INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THERE APPEARS TO BE NO detailed description of energy substrate fluxes, and particularly of plasma glucose kinetics, during prolonged running, except for the study by Hall et al. (11), which has only been presented as a preliminary report. Similarly, most of the studies concerning exogenous carbohydrate oxidation during exercise have been conducted on cycle ergometer. In these studies, the oxidation rate of exogenous carbohydrates has been shown to increase with workload (19, 27, 29) and with the amount ingested (1, 14, 36). At high absolute and relative workloads and for large amounts of carbohydrates ingested, the oxidation rate of exogenous glucose plateaus around 1.0 g/min, providing ~25% of the energy yield (13, 14, 26, 36). When running exercise was used, the workload was low [40% maximal oxygen consumption (O_2 max)] (22) or the amount of glucose ingested was small (0.4-0.8 g/min) (27, 28). Accordingly, the oxidation rate of exogenous glucose remained in the lower range of values reported (0.3-0.7 g/min). In the study by Derman et al. (10), exogenous glucose oxidation was compared in response to running and cycling exercise at a high workload (3.6 l O₂/min; 80% O_2 max) with ingestion of a large amount of glucose (2.4 g/min). The oxidation rate of exogenous glucose was lower during running than cycling exercise (0.230 and 0.345 g/min, respectively), providing <10% of the total energy yield. However, because of the high workload, the exercise time was comparatively short (cycling: 96 min; running: 63 min). Delays between glucose ingestion, absorption, and oxidation on one hand, and between production of labeled CO₂ in tissues and at the mouth (23) on the other hand, could explain the low oxidation rate observed in this study, and the lower oxidation rate observed during running compared with cycling exercise. In addition, gastric emptying and glucose absorption could be impaired during running exercise, which could be associated with gastrointestinal discomfort when carbohydrates are ingested (9, 31, 32). This could, in turn, reduce the availability and oxidation of exogenous carbohydrates.

The purpose of the present study was to verify this hypothesis and to describe the oxidation rate of a large amount of exogenous glucose during running exercise at a workload chosen to lead to exhaustion in ~2 h in a laboratory setting. The ingested glucose was artificially labeled with ¹³C to compute its oxidation rate from the production of ¹³CO₂ (¹³CO₂) at the mouth. In addition, oxidation of plasma glucose was computed from ¹³CO₂ and ¹³C/¹²C in plasma glucose, liver glucose output was estimated by difference between plasma and exogenous glucose oxidation, and the oxidation of muscle glycogen was computed from the difference between total glucose and plasma glucose oxidation (10, 26).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. Six active and healthy male subjects gave their written consent to participate in this study, which was approved by the Institutional Board on ethics on the use of human in research. Subject mean age, body mass (BM), height, and O_2 max on treadmill were 26.3 ± 2.8 yr, 70.5 ± 3.3 kg, 174.9 ± 3.7 cm, and 60.0 ± 2.2 ml · kg¹ · min¹, respectively (means ± SE). Subjects were running an average of ~40 km/wk, but none trained for competing in athletic events at a high level. All the subjects had a normal fasting plasma glucose concentration (4.7 ± 0.3 mmol/l). During the 3 days preceding each experiment, subjects refrained from exercising and drinking alcohol. They also avoided ingesting foods containing carbohydrates with a high ¹³C content (e.g., corn, sugar cane), which might modify background ¹³C enrichment of expired CO₂ (16). Compliance with this diet was confirmed by the low background ¹³C enrichment of expired CO₂, which was similar in the two experimental situations (Table 1).

View this table: [in this window] [in a new window]   Table 1.   O2, CO2, 13C/12C in expired CO2, and urea excreted in the placebo and glucose trials

Experimental protocol. O_2 max and experimental running speed at zero slope on treadmill (Quinton Q45, Bothell, WA) were determined for each subject during a preliminary test session by using open-circuit spirometry (1100 medical gas analyzer, Marquette Electronics, Milwaukee, WI). Then, all subjects performed, at a 7-day interval, two 120-min exercises at zero slope on the treadmill at a speed corresponding to 65% of the maximal running speed reached at O_2 max [11.4 ± 0.2 km/h; oxygen production (O₂) = 2,910 ± 97 ml O₂/min or 69 ± 2% O_2 max]. Each test was performed at 22 ± 1°C between 9:00 and 11:00 AM after an overnight fast, and a standardized breakfast was taken 2 h before the beginning of exercise (~500 kcal; ~48% carbohydrates, ~35% lipids, ~17% proteins). On the day before each experimental session, the evening meal was standardized and taken between 7:00 and 8:00 PM (~1,200 kcal; ~55% carbohydrates, ~25% lipids, ~20% proteins).

Measures and computations. Observations were made at rest before ingestion of the first dose of the solution, just before the beginning of exercise, and every 20 min during the exercise period. Total glucose and lipid oxidations were computed from indirect respiratory calorimetry corrected for protein oxidation. For this purpose O₂ and carbon dioxide production (CO₂) were measured by using open-circuit spirometry (5-min collection period), and urea production was estimated over the exercise period. Urea excretion over the 120 min of exercise was estimated from its concentration in urine (203 ± 48 and 355 ± 44 mmol/l in the placebo and glucose trials, respectively) and sweat (12.4 ± 1.8 and 12.9 ± 1.6 mmol/l), and from urine (0.36 ± 0.10 and 0.17 ± 0.03 liters) and sweat loss (2.75 ± 0.20 and 2.30 ± 0.21 liters) (3). Sweat loss was estimated from changes in BM, taking into account fluid intake, mass loss through CO₂, and water loss from the lungs (17). For the measurement of ¹³C/¹²C in expired CO₂, 80-ml samples of expired gas were collected in vacutainers (Becton Dickinson, Franklin Lakes, NJ). Finally, 15-ml blood samples were withdrawn by venopuncture at rest before the first ingestion of the solution and at minutes 60 and 120 during the exercise period for the measurement of plasma glucose, insulin, lactate, and free fatty acid concentrations, and for the determination of ¹³C/¹²C in plasma glucose. Plasma, urine, and sweat samples were stored at 80°C until analysis.

(1)

(2)

(3)

(4)

Statistics. Data are presented as means ± SE. The main effects of time and treatment (placebo and glucose ingestion) as well as time-ingestion interactions were tested by repeated-measure ANOVA (Statistica package). Newman-Keuls post hoc test was used to identify the location of significant differences (P  0.05) when ANOVA yielded a significant F ratio.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Urea excretion, O₂, and CO₂ during the exercise period were not significantly different in the placebo and glucose trials (Table 1). The ¹³C/¹²C in expired CO₂ at rest before ingestion of the solution was also similar in the placebo and glucose trials. In response to exercise in the placebo trial, a small (~1 [¹³C]PDB-1) but significant increase in ¹³C/¹²C in expired CO₂ was observed, as regularly reported (25). A much higher increase (10.3  [¹³C]PDB-1) was observed when [¹³C]glucose was ingested and oxidized (Table 1).

Table 2 presents the overall substrate utilization over the last 80 min of exercise. Protein oxidation, which was not significantly different in the placebo and control trials, provided, respectively, 4.9 and 4.2% of the total energy yield. When the placebo was ingested, total glucose and fat oxidation were, respectively, lower and higher between minutes 80 and 120 than between minutes 40 and 80. Glucose ingestion, respectively, increased and decreased total glucose and fat oxidation between minutes 80 and 120. This was because of the oxidation of exogenous glucose, which significantly increased from minutes 80 and 120. Over the last 80 min of exercise, 81.6 ± 6.6 g of exogenous glucose were oxidized, or 33% of the total amount ingested. In contrast, total endogenous glucose oxidation was significantly reduced when glucose was ingested and oxidized (184.6 ± 19.5 vs. 116.7 ± 18.4 g, providing 60 and 39% of the energy yield, respectively).

Figure 1 shows the rate of endogenous and exogenous glucose oxidation computed in the glucose trial. Changes in plasma glucose ¹³C/¹²C (21.02 ± 0.40, 1.30 ± 2.44, and 0.88 ± 1.70  [¹³C]PDB-1 before glucose ingestion, and at minutes 60 and 120 during the exercise, respectively) indicated that 70 ± 6 and 72 ± 7% of plasma glucose originated from exogenous glucose at minutes 60 and 120, respectively. Plasma glucose oxidation significantly increased from minutes 60 to 120 (from 1.30 ± 0.13 to 1.69 ± 0.16 g/min) because of both the increased oxidation rate of exogenous glucose (0.92 ± 0.06 g/min) and the glucose released from the liver (0.38 ± 0.09 and 0.47 ± 0.08 g/min), whereas the oxidation rate of muscle glycogen significantly decreased (from 1.28 ± 0.19 to 0.57 ± 0.11 g glucose/min). At the end of the first hour of exercise, muscle glycogen provided 49.7% of the glucose oxidized vs. 50.3% from plasma glucose (14.8% from the liver and 35.5% from the glucose ingested). At the end of the exercise period, plasma glucose provided 74.5% of the glucose oxidized (20.8% from the liver and 53.8% from glucose ingested) vs. only 25.5% from muscle glycogen.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Oxidation rates of glucose from blood, muscle, and a total value (A), and from blood, liver, and an exogenous source (B) at minutes 60 and 120 during the exercise period in the glucose trial. Values are means ± SE. aSignificantly different from minute 60 (P < 0.05).

Plasma glucose, insulin, free fatty acid, and lactate concentrations at rest and at the end of the hours 1 and 2 of exercise are presented in Fig. 2. No significant increase in plasma lactate concentration was observed in any of the two trials. In contrast, glucose ingestion resulted in a higher plasma glucose level at the end of exercise, and blunted the decrease in plasma insulin, and the rise in plasma free fatty acid concentration observed in response to exercise with placebo ingestion.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Changes in plasma glucose, insulin, free fatty acid, and lactate concentration. Values are means ± SE. aSignificantly different from rest (P < 0.05). bSignificantly different from placebo (P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Long distance runners often experience gastrointestinal discomfort when fed carbohydrates, possibly due to impairment in gastric emptying and/or intestinal absorption (9, 31, 32). This could reduce the availability and oxidation of exogenous carbohydrates. However, Rehrer et al. (30) did not report any major difference in gastric emptying in running and cycling at 70% O_2 max (3.15 l O₂/min) with ingestion of carbohydrates. In the present study, with an ingestion rate within the same range of that used by Wagenmakers et al. (36) (1.8-2.0 g/min) during cycling exercise at a similar workload (~3.0 l O₂/min; 65-70% O_2 max), the oxidation rate of exogenous glucose over the last 40 min of exercise (1.12 g/min) was very similar to that reported by these authors (1.07 g/min). The smaller values reported by Derman et al. (10) during both cycling and running exercise (0.54 and 0.40 g/min, respectively) with a high ingestion rate (2.4 g/min) could be due to the shorter period of exercise (cycling: 96 min; running: 63 min) and the much higher relative workload sustained (80% O_2 max). In support of this hypothesis, McConell et al. (21) have shown that, during a 68-min cycling exercise at 83% O_2 max, only a small portion of ingested glucose entered the plasma (22 vs. 84 g ingested), possibly due to a slow rate of gastric emptying and intestinal absorption at such a high exercise intensity.

Most studies of plasma glucose oxidation, with or without glucose ingestion, have been performed during cycling exercise (2, 5, 12, 14, 20, 21, 26). During running exercise, with ingestion of 150 g of glucose, Derman et al. (10), using [¹⁴C]glucose, have shown that plasma glucose oxidation was 0.83 g/min (0.40 and 0.43 g/min from the liver and from exogenous glucose, respectively) at the end of a 63-min exercise period at 80% O_2 max, providing 23% of the energy yield. A slightly higher value (1.4 g/min) can be computed from data reported by Tsintzas et al. (35) during a 104-min running exercise at 70% O_2 max, with ingestion of 0.6 g of glucose/min; total glucose oxidation computed from indirect calorimetry averaged 3.1 g/min, whereas muscle glycogen oxidation averaged 1.7 g/min (2.5 mmol · min¹ · kg¹ × 3.8 kg of working muscles, dry weight). In the study by Hall et al. (11), the rate of plasma glucose disappearance measured by using [¹⁴C]glucose during a 150- to 165-min running exercise at ~76% O_2 max averaged only 0.45 g/min, providing 12.2% of the energy yield, but no carbohydrates were ingested.

In the present experiment, with a large amount of glucose ingested (~250 g), glucose release from the liver slightly increased from minutes 60 to 120 (0.38 and 0.47 g/min, respectively) and was well in accordance with data reported at similar workloads during running (11) and cycling exercise (2, 5, 14, 20, 21, 26). Also, as already reported during cycling exercise with carbohydrate ingestion (2, 5, 10, 14, 20, 21, 26), a much higher amount of plasma glucose originated from exogenous glucose (0.92 and 1.22 g/min at minutes 60 and 120, respectively, representing 70-72% of plasma glucose and 36-54% of total glucose oxidation). The rate of plasma glucose oxidation, which was well above 1 g/min at minute 60, reached 1.69 g/min at the end of exercise. These oxidation rates of plasma glucose observed with ingestion of ~2 g of glucose/min, and with plasma glucose concentration between 7.5 and 8 mmol/l over the last hour of exercise, are in line with those reported by Hawley et al. (12) and Weltan et al. (37). In these studies, with plasma glucose increased to ~9 to 9.7 mmol/l by using glucose infusion (up to ~2.5 to 3 g/min) during prolonged exercise (120-145 min) at 70% O_2 max, the oxidation rate of plasma glucose peaked at ~1.75 to 2 g/min. As discussed by McConell et al. (21) the very high oxidation rates of plasma glucose during exercise at 50-70% O_2 max, with large amounts of glucose administered, could be due to the high plasma glucose and insulin concentrations, and low plasma free fatty acid concentration, combined to the effect of muscle contraction per se on glucose uptake. In addition, as discussed by Tsintzas and Williams (33), glucose ingestion appears to reduce muscle glycogen utilization during running but not during cycling exercise. This could be associated with a higher rate of plasma glucose oxidation during running than cycling and could explain the comparatively high value observed in the present study.

The oxidation rate of muscle glycogen can be estimated from changes in muscle glycogen content (7, 34, 35) or could be computed by difference between the total oxidation rate of glucose (computed by using indirect respiratory calorimetry) and of plasma glucose (computed by using tracer techniques) (2, 5, 10, 14, 20, 21, 26). Using this method, Derman et al. (10) and McConell et al. (21) have shown at 80-83% O_2 max that the muscle relies much more on muscle glycogen (3.1-3.4 g/min providing 70-82% of the energy yield) than on plasma glucose oxidation (0.8-0.9 g/min). In contrast, for the percentage of O_2 max ranging between 50 and 70%, the oxidation rate of plasma glucose during cycling exercise is much higher, particularly if exogenous glucose is administered, and the oxidation rate of muscle glycogen ranges only between 0.6 and 2.1 g/min, providing 20-58% of the energy yield (2, 5, 14, 20, 26). Data from the present experiment, obtained during running exercise, are in accordance with these previous findings at similar absolute and relative workloads. The oxidation rate of muscle glycogen, which was 1.28 ± 0.19 g of glucose/min at minute 60, providing 33.7% of the energy yield, was reduced to 0.57 ± 0.16 g of glucose/min at the end of the exercise period, providing only 15.5% of the energy yield. These values are lower than the average oxidation rate of muscle glycogen estimated by Tsintzas et al. (1.7 g/min providing 37% of the energy yield) (35). This could stem from the fact that the amount of glucose administered was much higher than in the study by Tsintzas et al. (247 vs. 65 g) (35) and that plasma glucose oxidation was also higher (1.7 vs. 1.4 g/min at minute 120) with, possibly, a larger muscle glycogen sparing.

In summary, results from the present experiment suggest that during prolonged running exercise at ~70% O_2 max with ingestion of large amounts of glucose (~2 g/min), as already shown during cycling exercise, exogenous glucose and plasma glucose, and not muscle glycogen, are the main sources of glucose for oxidation, particularly at the end of exercise. During cycling exercise, the large contribution of exogenous glucose oxidation to the energy yield reduces liver glucose output (2, 14, 20, 21) but does not consistently reduce muscle glycogen utilization (Refs. 2, 14, 21; see Ref. 33 for review). When carbohydrates are ingested during running exercise, no data are available concerning possible changes in liver glucose output, but consistent data indicate that muscle glycogen utilization is reduced (Refs. 34, 35; see Ref. 33 for review). In the present experiment, the large oxidation rate of exogenous glucose was associated with a marked 37% reduction of endogenous glucose oxidation (68 g between minutes 40 and 120). However, it cannot be ascertained whether this was because of a reduction in muscle and/or liver glycogen utilization.