INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCSSION REFERENCES

BLOOD GLUCOSE IS AN IMPORTANT fuel for contracting muscle. Studies in several species have demonstrated that glucose uptake [rate of dissapearance (R_d)] by muscle increases during physical exercise (6). Therefore, an increase in hepatic glucose production (HGP) is required to meet this demand for blood glucose and to avoid hypoglycemia. Indeed, studies in both humans (22) and dogs (3) during moderate-intensity exercise have demonstrated that the dynamics of HGP are similar to those of the increase in glucose R_d, such that plasma glucose concentrations are maintained within a narrow range. Furthermore, glucose infusion in an amount corresponding to the exercise-induced increase in HGP has been demonstrated to largely abolish endogenous glucose production [rate of appearance (R_a)] in these species during moderate exercise (3, 22). These findings suggest that metabolic feedback mechanisms associated with an imbalance between glucose supply and demand stimulate the exercise-induced increase in HGP. During more strenuous exercise [>70% of maximum oxygen uptake (O_{2 max})], however, HGP greatly exceeds glucose R_d, and plasma glucose concentrations increase, suggesting that other mechanisms contribute to the mobilization of liver glucose in these circumstances. In support of this hypothesis, some (26, 40), but not all (21), studies have demonstrated that infusion of glucose only partially attenuates HGP during heavy exercise. During exercise at these higher workloads, neural feed-forward mechanisms, rather than metabolic feedback, may control the increase in HGP.

In horses, unlike humans and dogs, plasma glucose concentrations increase (~2-4 mM) even during moderate-intensity exercise (35-50% O_{2 max}) (13, 33). This finding indicates a mismatch between glucose R_a and R_d in the horse during moderate exercise and suggests that neural feed-forward mechanisms may, in part, control liver glucose mobilization. However, because there are no reports of whole body glucose turnover in the horse during sustained exertion, no information exists concerning glucoregulatory mechanisms in this species. We hypothesized that, if feed-forward mechanisms control the glucose R_a response to exercise independent of glucose feedback, an intravenous glucose infusion would not suppress HGP under such circumstances. Therefore, the first objective of this study was to determine the effects of an intravenous glucose infusion on HGP in horses during prolonged, low-intensity (~35% O_{2 max}) exercise.

In humans, it is well established that provision of exogenous glucose (intravenous infusion or glucose ingestion) influences metabolism during exercise (2, 8). Specifically, an increase in blood glucose availability augments glucose R_d by muscle and, during prolonged mild-to-moderate-intensity exercise (i.e., 30-70% O_{2 max}), attenuates the decrease in carbohydrate oxidation (CHO_ox) and enhances endurance performance. However, most (9, 10, 17), but not all (36, 37), studies in humans have reported that net muscle glycogen utilization is not altered by carbohydrate ingestion. In horses during moderate-intensity exercise, intravenous glucose infusion (~2 g/min) prolonged running performance by ~14% relative to the control treatment (13). Thus, similar to other species, carbohydrate administration exerts an ergogenic effect in horses during prolonged exercise. However, the effects of an increase in blood glucose availability on tracer-determined glucose uptake, muscle glycogen utilization, and CHO_ox have not been simultaneously studied in the horse. Therefore, the second objective of this study was to determine the effects of intravenous glucose infusion on glucose R_d, whole body CHO_ox, and muscle glycogen utilization in horses during 90 min of exercise at 35% O_{2 max}. We hypothesized that an increase in glucose availability would increase R_d and attenuate the decline in CHO_ox during such exercise but would not alter muscle glycogen utilization.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCSSION REFERENCES

All animal experiments were conducted after approval by the Institutional Laboratory Animal Care and Use Committee of The Ohio State University and were performed in compliance with their guidelines and recommendations.

Experimental Design

1

1

Horses

Preliminary Testing

1

1

Experimental Protocol

1

1

1

1

Respiratory Gas-Exchange Measurements

1

1

Sampling Procedures

Muscle biopsy samples were collected percutaneously from the middle gluteal muscle by using the needle-biopsy technique. Muscle biopsies were obtained 5 min before commencement of exercise and within 3 min of cessation of exercise. Muscle samples were immediately placed in liquid nitrogen and stored at 80°C until analysis.

Analytic Techniques

Plasma isotopic enrichment. For determination of [6,6-²H]glucose enrichment, plasma samples (0.7 ml) were deproteinized by adding 1.2 ml of 0.3 N Zn(SO)₄ and 1.2 ml of 0.3 N Ba(OH)₃. Each tube was then vortexed and incubated in an ice bath for 20 min. After centrifugation at 3,000 rpm for 20 min at 4°C, the supernatant was harvested and placed in 13 × 100-mm screw-cap borosilicate tubes. The water was removed from the tubes by vacuum centrifugation (Savant Instruments, Farmingham, NY). The dried samples were capped and stored at 4°C until further processing. The penta-acetate derivative of glucose was prepared by adding 100 µl of a 2:1 acetic anhydride and pyridine mixture to the dried sample. After a 60-min incubation at 60°C in a dry heating block, the reaction mixture was transferred to a clean 13 × 100-mm borosilicate tube. The penta-acetate derivative was partitioned by sequential addition of 1.5 ml of double-distilled water and 0.4 ml of methylene chloride. After gentle shaking, the tubes were centrifuged at 2,000 rpm for 10 min. The upper water phase was discarded, and the remaining methylene chloride phase was evaporated under N₂. Before injection into the gas chromatograph-mass spectrometer, the samples were dissolved in 50 µl of ethyl acetate. Standards of known isotopic enrichment were prepared in an identical fashion and were analyzed with each batch of samples.

Plasma biochemical analyses. Plasma glucose concentration was measured spectrophotometrically by using the hexokinase reaction with a commercial kit (Glucose-HK kit; Sigma Chemical, St. Louis, MO), and plasma lactate concentration was measured by using an automated lactate oxidase method (Sport 1500 lactate analyzer; Yellow Springs Instruments, Yellow Springs, OH). Plasma NEFA concentration was determined by using a commercial kit that employs an enzymatic colorimetric method (NEFA test kit; Wako Chemicals USA, Dallas, TX). Plasma glycerol concentration was measured by using an enzymatic spectrophotometric method [triglycerides kit 337A (without triglyceride hydrolysis step); Sigma Chemical]. Intra- and interassay coefficients of variation for these biochemical methods were <1.0 and 2.5%, respectively. Hematocrit was measured by the microhematocrit technique. Plasma total protein was measured by refractometry (Cambridge Instruments, Buffalo, NY). All samples were analyzed in duplicate.

Plasma hormone analyses. Plasma Epi and NE concentrations were determined by HPLC using electrochemical detection (25). Serum immunoreactive insulin (IRI) was determined in duplicate by use of a commercially available radioimmunoassay (insulin kit, Coat-a-Count Diagnostics, Los Angeles, CA) that has been validated for horse blood (31). Intra- and interassay coefficients of variation were 6.0 ± 1.5 and 11.5 ± 2.1%, respectively. Plasma immunoreactive glucagon (IRG) was determined in duplicate by use of a commercially available radioimmunoassay (glucagon kit, Coat-a-Count Diagnostics). Pooled equine plasma was used to partially validate the assay for horse plasma. Specificity for equine glucagon was demonstrated by dilutional parallelism between standard solutions and serial dilutions of endogenous glucagon in equine plasma [correlation coefficient (r) = 0.987]. Accuracy was demonstrated by adding porcine glucagon to equine plasma in concentrations from 20 to 285 pmol/l. Linear regression of the recovery curve showed an r of 0.9929. The intra-assay precision for 12 replicates (6 duplicates) of equine plasma with a mean concentration of 31 and 75 pmol/l was 7.8 and 5.2%, respectively. The interassay coefficient of variation for the same samples was 13.1 and 14.8%, respectively. For the insulin and glucagon radioimmunoassays, analysis of experimental samples was completed in a single analytic session.

Muscle glycogen. Muscle samples were weighed and subsequently freeze-dried. The freeze-dried samples were reweighed and dissected free of any blood and connective tissue. For each sample, duplicate pieces of muscle were analyzed. The samples were powdered, extracted, and analyzed for muscle glycogen concentration (as glucosyl units) according to the procedure of Passoneau and Lauderdale (29). Intra-assay (12 replicates of a single sample) and interassay (6 replicates) coefficients of variation for this assay were 2.5 and 4.1%, respectively.

Calculations of glucose kinetics. Glucose R_a and R_d (= tissue uptake) at rest were calculated by using the steady-state tracer dilution equation (41)
where F is the infusion rate of the isotope (in µmol · kg¹ · min¹); IE_i and IE_p are the stable isotopic enrichment of the infusate and plasma, respectively; and 1 accounts for the tracer's contribution to the turnover rate of the substrate (41). The rate of infusion was calculated by multiplying the infusion pump rate by the concentration of glucose in the infusate. During exercise, R_a and R_d were calculated by using the non-steady-state equation of Steele (35). This equation is modified for use with stable isotopes as the amount of tracer infused is no longer negligible
and

where V_d is the effective volume of distribution, E is the plasma isotopic enrichment, C_m is the measured plasma concentration of the tracee, and dE/dt and dC_m/dt are maximum rates of change in enrichment and glucose concentration, respectively, as a function of time. By using this fixed, one-compartment model of Steele, it is assumed that 1) the apparent glucose space is 25% of body weight, and 2) 65% of this space represents the rapidly mixing portion of the glucose pool. Therefore, the effective V_d for glucose was assumed to be 162 ml/kg.

Rates of energy expenditure and whole-body substrate oxidation. Total energy expenditure (TEE) and absolute rates of CHO_ox and lipid oxidation were calculated using the following equations (16)

(1)

(2)

(3)

where O₂ and CO₂ are in liters per minute, and it was assumed that protein oxidation made negligible contribution to O₂ and CO₂ and that these equations, developed for use in humans, are also applicable to the horse. The calculated values were based on respiratory gas-exchange values averaged over the first 5 min of exercise and 15-min intervals thereafter. CHO_ox in grams per minute was converted to micromoles per kilogram per minute by dividing the molecular weight of glucose (180) and the horse's body weight. Similarly, rates of fat oxidation were converted to micromoles per kilogram per minute by dividing by the molecular weight of palmitate (259) and the horse's body weight.

Statistical analyses. Data are presented as means ± SE. The data were analyzed by using a two-way analysis of variance for repeated measures, with treatment (Con vs. Glu) and time as independent factors. Because the data for Epi and NE did not exhibit homogeneous variances, these data were subject to logarithmic transformation before ANOVA. Percent data (relative contributions by different substrates to TEE) were subject to arcsine transformation before ANOVA. The null hypothesis was rejected at  = 0.05 for the main effects (treatment and time) and  = 0.10 for the interaction. Significant differences identified by ANOVA were isolated by using the Student-Newman-Keuls post hoc test. Data for net muscle glycogen utilization and total CHO_ox were analyzed by paired t-test. The Sigmastat 2.0 software package (Jandel Scientific, San Rafael, CA) was used for statistical computations.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCSSION REFERENCES

Individual values for O_{2 max} ranged from 121 to 164 ml · kg¹ · min¹ (mean 136.3 ± 6.1 ml · kg¹ · min¹). Mean running speed during the exercise protocol was 4.4 ± 0.2 m/s, which corresponded to a relative workload of 34 ± 1.0% of O_{2 max}. O₂ during exercise was relatively constant and not significantly different between trials. However, whereas RER declined progressively throughout exercise in both trials, glucose infusion attenuated (interaction effect, P < 0.02) this decrease, and RER was significantly higher in Glu than in Con at 30, 75, and 90 min of exercise (Table 1).

Plasma Glucose Concentration and Kinetics

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Plasma glucose concentration (A) and isotopic enrichment (B) at rest and during 90 min of exercise at 34 ± 1% of maximum oxygen uptake with (glucose) and without (control) glucose infusion commencing at onset of exercise. Values are means ± SE for 6 horses. * Significantly different from control, P < 0.05.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Rate of plasma glucose appearance (total Ra; A) and hepatic glucose production (HGP; B) before and during 90 min of exercise at 34 ± 1% of maximum oxygen uptake under control and glucose infusion conditions. Values are means ± SE for 6 horses. * Significantly different from control, P < 0.05.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Whole body glucose uptake [glucose rate of disappearance (Rd); A] and metabolic clearance rate (MCR) (B) before and during 90 min of exercise at 34 ± 1% of maximum oxygen uptake under control and glucose infusion conditions. Values are means ± SE for 6 horses. * Significantly different from control, P < 0.05.

Total glucose R_a (HGP + exogenous glucose) was 60-70% higher in Glu than in Con throughout exercise (Fig. 2A). Endogenous glucose R_a (HGP) in Con increased progressively during exercise and reached a peak of 38.6 ± 2.7 µmol · kg¹ · min¹ after 90 min (Fig. 2B). Glucose infusion suppressed (P < 0.05) the exercise-associated increase in HGP. However, this suppression of HGP during exercise in Glu was not complete, and values for endogenous R_a increased from 8.4 ± 0.6 µmol · kg¹ · min¹ at rest to a peak of 25.8 ± 3.3 µmol · kg¹ · min¹ at 90 min (Fig. 2B).

In Con, the exercise-associated increase in glucose R_d was biphasic; R_d rose from 8.3 ± 0.9 to 29.1 ± 3.1 µmol · kg¹ · min¹ at 45 min and further increased between 60 and 90 min to a peak of 40.4 ± 2.9 µmol · kg¹ · min¹ (Fig. 3A). Glucose R_d was significantly higher in Glu than in Con throughout exercise (P < 0.001). Between 30 and 90 min of exercise, glucose R_d was approximately twofold higher in Glu than in Con (Fig. 3A). MCR was similar at rest in the two trials and increased to a similar extent during the initial 15 min of exercise. From 30 min until the end of exercise, MCR was significantly greater in Glu than in Con (Fig. 3B).

Plasma Hormone Concentrations

Hematocrit, Plasma Total Protein, Lactate, Glycerol, NEFA

Energy Expenditure and Substrate Utilization

1

1

1

1

View larger version (37K): [in this window] [in a new window]   Fig. 4.   Absolute (A) and relative (B) contributions of energy from different substrate sources for 3 periods during 90 min of exercise at 34 ± 1% of maximum oxygen uptake under control (Con) and glucose (Glu) infusion conditions. Values are means ± SE for 6 horses. CHO, carbohydrate. * Significantly different from control, P < 0.05.

Areas under the total CHO_ox and plasma glucose R_d vs. time curves during the two trials are shown in Fig. 5. Total CHO_ox was significantly higher in Glu (1,435 ± 65 g/90 min) than in Con (1,185 ± 55 g/90 min). Total calculated glucose R_d was also higher (P < 0.001) in Glu (375 ± 20 g/90 min) compared with Con (205 ± 15 g/90 min). However, muscle glycogen (plus lactate) oxidation, estimated from the difference between the total CHO_ox and total glucose R_d, was not significantly different between the two trials (Con: 980 ± 45 g/90 min, Glu: 1,060 ± 80 g/90 min; Fig. 5). Net muscle glycogen utilization, calculated as the difference in muscle glycogen concentration between pre- and post-exercise biopsy samples, was also similar in the Con (164 ± 31 mmol glucosyl units/kg dry muscle) and Glu (161 ± 24 mmol glucosyl units/kg dry muscle) trials (Fig. 6).

View larger version (49K): [in this window] [in a new window]   Fig. 5.   Total CHO oxidation during 90 min of exercise at 34 ± 1% of maximum oxygen uptake under control and glucose infusion conditions. Values are means ± SE for 6 horses. Muscle glycogen (plus lactate) oxidation was estimated from difference between total CHO oxidation and glucose Rd. * Total carbohydrate oxidation significantly greater than control, P < 0.05. ** Total calculated glucose Rd greater than control, P < 0.05.

View larger version (41K): [in this window] [in a new window]   Fig. 6.   Muscle glycogen concentration before (Pre-Ex) and after (Post-Ex) 90 min of exercise at 34 ± 1% of maximum oxygen uptake under control and glucose infusion conditions. dw, Dry weight. Values are means ± SE for 6 horses. * Significantly different from Pre-Ex, P < 0.05.

Estimates of the absolute and relative caloric contributions from plasma glucose, other carbohydrate sources (muscle glycogen plus lactate), and fat during the 0- to 30-min, 30- to 60-min, and 60- to 90-min periods of exercise are shown in Fig. 4. The higher CHO_ox in Glu than in Con (Table 1) reduced (P < 0.05) the percent contribution to total energy use by fat oxidation from between 43 and 68% in the Con trial to between 35 and 55% in the Glu trial (Fig. 4B). Conversely, the contribution by plasma glucose to energy use was significantly higher in Glu than in Con. In Glu, the percent contribution by plasma glucose to TEE increased from 12 ± 1.5% during the 0- to 30-min period to 20 ± 2.2% during the 60- to 90-min period. Corresponding values for the Con trial were 6.1 ± 1.2 and 11.2 ± 2.1% (Fig. 4B). In both trials, the contribution from muscle glycogen (and lactate) to TEE decreased from ~50-52% during the 0- to 30-min period to ~20-23% during the 60- to 90-min period. The contribution from muscle glycogen was not different between trials.

	DISCSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCSSION REFERENCES

The present study is the first in horses to determine the effects of an increase in glucose availability on the kinetics of endogenous glucose production and whole body R_d during exercise. In addition, this study is the first, to our knowledge, to quantify the effects of an increase in glucose supply on the relative contributions from different substrate sources to TEE in horses during low-intensity exercise. The principal findings were that 1) an infusion of glucose at a rate that approximated the average endogenous R_a measured in Con trials only partially suppressed hepatic glucose output during exercise; 2) the increase in glucose availability as a result of glucose infusion (~3 g/min) resulted in twofold increases in whole body glucose R_d and the estimated contribution by plasma glucose to TEE; and 3) glucose infusion did not affect muscle glycogen utilization but attenuated the decrease in CHO_ox and increase in fat oxidation measured during exercise in Con trials.

Critique of Methods

When exogenous (unlabeled) glucose is infused, endogenous glucose R_a is calculated as the difference between the total glucose R_a measured by using the tracer and the known rate of glucose infusion. However, as discussed by Coggan et al. (7), with use of stable isotopically labeled tracers, errors are introduced into this calculation when the isotopomer distribution of the exogenous glucose (R_exo) differs from that of the endogenous glucose (R_o). If these isotopomer distributions differ, the exogenous unlabeled glucose is either enriched (R_exo > R_o) or depleted (R_exo < R_o) compared with the endogenous tracee, and the rate of tracer infusion must be corrected to account for the effects of the exogenous glucose on the "enrichment" of the glucose pool. In the present study, for each trial in which unlabeled glucose was infused, the exogenous substrate was "enriched" compared with the m/z 202/200 ion-abundance ratio of the horse's endogenous glucose. However, these differences were very small (~1.0-1.5%), resulting in only minor adjustments to the tracer and exogenous tracee infusion rates.

Glucose Kinetics

Inasmuch as exogenous glucose only partially suppressed HGP, the results of the present study differ from those of previous studies in humans and animals. In humans and dogs during moderate-intensity exercise, an intravenous glucose infusion that matches the endogenous R_a response measured during Con trials completely inhibits the exercise-associated increment in HGP (3, 21, 22). Although the reasons for this species variation cannot be determined from our data, it is possible that greater sympathoadrenal activation in the horse during low-intensity exercise plays a role in the feed-forward regulation of hepatic glucose output. After 100 min of moderate-intensity (2-fold increase in heart rate) exercise in dogs, plasma NE and Epi concentrations were approxiamtely twofold higher than at rest (3). Similarly, in trained human athletes exercising at ~50% of O_{2 max}, plasma NE and Epi were only two- to threefold higher compared with preexercise values (12). In contrast, in the Con trial of the present study, peak values for plasma NE and Epi were five- and ninefold higher, respectively, compared with resting concentrations (Table 2). The more striking increase in plasma catecholamines in the horse likely reflects a greater degree of sympathoadrenal activation compared with responses measured in dogs and humans at similar relative exercise intensities. Increases in circulating catecholamines, and possibly direct sympathetic neural activation, can enhance hepatic glucose mobilization during exercise, although species differences exist (38). Elucidation of the role of sympathoadrenal mechanisms in stimulation of HGP in the exercising horse awaits further investigation.

Several mechanisms may have contributed to the attenuation of HGP during Glu. In particular, glucose infusion altered the plasma concentrations of hormones that, in other species, are important for glucoregulation. Serum IRI concentrations were higher in Glu, whereas the increase in plasma IRG was attenuated by glucose infusion. Therefore, the insulin-to-glucagon molar ratio was higher during exercise in Glu (Table 2). In dogs (38), changes in these pancreatic hormones, specifically an increase in plasma glucagon with a concomitant decrease in plasma insulin, play an important role in regulation of the exercise-induced increment in hepatic glycogenolysis. Assuming such changes affect hepatic glucose output in horses during exercise, the higher serum IRI and lower plasma IRG concentrations could account for a reduction in HGP. Plasma Epi concentrations also were lower in Glu than in Con during exercise (Table 2). However, this difference was small and did not reach statistical significance until the end of exercise. Nonetheless, the reduced Epi concentrations may have contributed, in part, to the decrease in hepatic glucose output. Finally, the elevated plasma glucose concentration, independent of changes in glucoregulatory hormones, may have reduced HGP directly. In rats, increases in plasma glucose directly lower hepatic glycogenolysis and HGP by decreasing glycogen phosphorylase and increasing glucokinase activities (34). However, the progressive increase in HGP during the final 45 min of exercise in Glu, when mean plasma glucose concentrations were >7.5-8.0 mmol/l (Fig. 1A), argues against a significant role for this mechanism in the present study.

Given the substantial endogenous R_a response in Glu, total glucose supply (the sum of exogenous glucose and HGP; Fig. 2A) was approximately twofold higher compared with that in the Con trial. Furthermore, by a similar magnitude, whole body glucose R_d was higher in Glu than Con throughout exercise (Fig. 3A). As discussed in MATERIALS AND METHODS, it is assumed that >90% of tracer-determined whole body glucose R_d reflects glucose uptake by muscle. Because we did not collect urine from the horses of the present study, we cannot eliminate the possibility that at least some of the R_d reflected loss of glucose in urine. However, in horses the renal threshold for glucose is ~10.5-11 mmol/l (30), although it is possible that this threshold is lower during exercise. Nevertheless, the contribution of renal losses to glucose R_d, if any, was likely to be small.

A number of factors regulate glucose uptake by muscle, including glucose delivery to muscle, membrane glucose transport, and intracellular metabolism. Glucose delivery is a function of muscle blood flow and prevailing glucose concentrations. In exercising dogs, tracer-determined whole body and leg R_d are augmented after an increase in blood glucose availability (42). Similarly, carbohydrate ingestion or intravenous glucose infusion that elevates plasma glucose concentrations has been demonstrated to increase leg (1) and whole body (10, 21, 27) glucose disposal in humans during moderate-intensity exercise. Therefore, part of the increment in glucose uptake in Glu can be attributed to an increase in blood glucose concentration. However, MCR (R_d/glucose concentration) was significantly higher in Glu than in Con between 30 and 90 min of exercise (Fig. 3B), suggesting that other factors may have contributed to the increase in glucose R_d. Studies in humans (39) and dogs (42) have demonstrated that muscle contractions and insulin have synergistic effects on glucose R_d by muscle during exercise. In addition, a suppression of circulating NEFA availability has been shown to increase whole body and limb glucose R_d significantly in dogs during exercise (4). In the present study, serum IRI was higher during exercise in Glu (Table 2), whereas glucose infusion attenuated the exercise-associated increase in free fatty acids such that plasma NEFA concentrations were significantly lower in Glu than in Con between 45 and 90 min of exercise. Therefore, it is possible that a combination of higher serum IRI and lower plasma NEFA in Glu contributed to the increment in glucose R_d and MCR.

Another factor that may have contributed to the increased glucose R_d and MCR in Glu was the suppression in plasma Epi concentrations (Table 2). In human subjects during low-intensity exercise (~25% O_{2 max}), an infusion of Epi that results in a modest increase in plasma Epi concentration (control ~0.6 nmol/l vs. Epi infusion ~2.0 nmol/l) is associated with an ~20% reduction in glucose clearance (28). Similarly, in resting human subjects, physiological increases in plasma Epi constrain glucose R_d and clearance (32). Therefore, it is possible that the lower plasma Epi concentrations in Glu during the last 30 min of exercise, in part, contributed to the increase in glucose R_d and MCR.

Substrate Utilization

There are little published data in horses on the effects of an acute increase in glucose availability on muscle glycogen utilization during exercise. However, in accord with the findings of the present study, Farris et al. (13) reported that an intravenous glucose infusion that elevated plasma glucose concentrations to ~9-10 mmol/l had no effect on glycogen utilization in the middle gluteal muscle of horses during exercise at ~50% O_{2 max}. In humans during prolonged cycling exercise, carbohydrate feedings that maintain blood glucose at euglycemic levels do not alter net muscle glycogen utilization (9). Similarly, a glucose-infusion protocol that maintains plasma glucose concentrations at ~10 mmol/l does not affect the net rate of muscle glycogen utilization in trained men during intense exercise (~73% O_{2 max}) (10). In contrast, carbohydrate ingestion by trained men during running at 70% O_{2 max} was associated with an ~28% reduction in net glycogen breakdown in type I fibers of vastus lateralis muscle (37).

If it is assumed that all of the glucose R_d was oxidized, plasma glucose supplied 26 ± 2% of the total CHO_ox and 16.5 ± 2% of the energy yield during exercise in Glu. In contrast, corresponding values for the contribution by plasma glucose in Con were 17 ± 2% of the total CHO_ox and 8.5 ± 1% of the energy yield. In the Glu trial, calculations of total glucose R_a indicate that ~430-440 g of glucose were available during the 90 min of exercise (280 ± 10 g exogenous glucose plus 155 ± 12 g from endogenous production). Therefore, there was a relatively small disparity between the quantity of glucose available and the calculated amount utilized during exercise (375 ± 20 g/90 min). On the basis of the glucose distribution volume used in the kinetic calculations, ~220 mmol (40 g) of glucose would be needed to increase the plasma glucose concentration by ~3 mmol/l (Fig. 1A).

The higher total CHO_ox in Glu was accompanied by a significant attenuation of fat oxidation (Table 1). Furthermore, plasma NEFA concentrations were significantly lower in Glu than in Con between 45 and 90 min of exercise (Table 3). Given the higher serum IRI, it is possible that the lower NEFA concentration levels were due to a reduction in lipolysis (19). Recent studies in humans during low- and moderate-intensity exercise have shown that increased blood glucose availability suppresses fat utilization by inhibition of both fat mobilization (i.e., lipolysis in adipose tissue) and fat oxidation within muscle (11, 19). Horowitz and colleagues (19) have shown that relatively small increases in plasma insulin (~10-30 µU/ml) associated with fructose or glucose ingestion reduce whole body lipolysis and fat oxidation by ~40-50% during a subsequent bout of moderate-intensity exercise (44 ± 2% peak O₂). Furthermore, this group has demonstrated that the increase in whole body glucose disposal associated with preexercise glucose ingestion is accompanied by inhibition of long-chain fatty acid oxidation (11). It is likely that similar mechanisms accounted for the reduction in fat oxidation measured in the Glu trial.

In summary, this study has demonstrated that an intravenous glucose infusion at a rate that approximated the average endogenous R_a measured in Con trials only partially suppressed hepatic glucose output in horses during low-intensity (~35% O_{2 max}) exercise. This finding suggests that, in addition to blood-borne metabolic feedback signals, feed-forward mechanisms may regulate HGP in horses during moderate exercise. The hyperglycemia and hyperinsulinemia associated with glucose infusion resulted in an approximately twofold increases in whole body glucose uptake and the estimated contribution of plasma glucose to energy expenditure. Glucose infusion did not affect muscle glycogen utilization. However, the increase in plasma glucose uptake and utilization attenuated the decrease in CHO_ox measured in the Con trial.