INTRODUCTION

Top Abstract Introduction References

METABOLIC ADAPTATION to training includes a shift toward an increase in fat oxidation (2, 14) and a decrease in plasma glucose turnover and oxidation in response to prolonged exercise (4, 5, 24, 32). However, training has also been shown to improve whole body glucose uptake during oral and intravenous glucose tolerance test, and euglycemic-hyperinsulinemic clamps, because of an increased insulin sensitivity and responsiveness (see Ref. 8 for review). In addition, non-insulin-mediated glucose transport in muscle during exercise is larger in trained subjects (39). Thus, despite a lower dependence on carbohydrate oxidation during prolonged exercise, trained subjects could have a greater ability to oxidize plasma glucose and, therefore, exogenous glucose when it is supplied at rest as well as during exercise.

Only two groups of authors have described the effect of training state on exogenous [¹³C]glucose oxidation at rest (18) and during exercise (16, 17). After a 6-wk training program at 30-40% maximal oxygen consumption (O_{2 max}), Krzentowski et al. (17, 18) found no difference in the oxidation rate of 100 g of [¹³C]glucose ingested at rest (35.9 vs. 37.4 g over 7 h) (18) but a significant 10% increase during exercise (~28 vs. ~31 g over 90 min) (17). In the recent cross-sectional study by Jeukendrup et al. (16), over the last hour of a 120-min exercise period with ingestion of 100 g of glucose, although the difference did not reach statistical significance, the amount of exogenous glucose oxidized was also 10% higher in trained (50 g) than in sedentary subjects (45 g). The effect of training on the oxidation rate of exogenous glucose during exercise for given absolute and relative workloads remains difficult to ascertain, however. In the study by Krzentowski et al. (17), the absolute workload was kept similar before and after training [oxygen uptake (O₂) ~1.3 l/min, corresponding to 41 and 32% O_{2 max} before and after training, respectively]. In the study by Jeukendrup et al. (16), trained and sedentary subjects both exercised at 50% of their respective maximal mechanical power output. As a consequence, both the relative (50 vs. 63% O_{2 max}) and absolute workloads (146 vs. 207 W or 2.2 vs. 2.6 l O₂/min) were different.

The purpose of the present experiment was, thus, to compare the oxidation rate of exogenous glucose ingested at rest as well as during prolonged exercise performed at the same relative and absolute workloads, in trained and sedentary subjects, by using ¹³C-labeling. In addition, plasma glucose ¹³C/¹²C was used to compute the oxidation of plasma glucose. The oxidation of glucose released from the liver was computed by difference between plasma and exogenous glucose oxidation, and the oxidation of glucose released from glycogen in periphery (mainly muscle glycogen during exercise) was computed by difference between total glucose oxidation and plasma glucose oxidation.

	METHODS

Subjects. The experiment was conducted on six well-trained endurance cyclists and six healthy but sedentary male subjects with a normal fasting plasma glucose concentration (4.48 ± 0.22 mmol/l; n = 12) (Table 1). Sedentary subjects did not regularly take part in any physical activity and had no history of endurance training. All subjects gave their informed written consent to participate in the study, wich was approved by the Institutional Board on the Use of Human Subjects in Research. None of the subjects were smokers, heavy drinkers, under medication, or using recreational drugs.

Assessment at exercise. O_{2 max} and experimental workloads on cycle ergometer (Ergomeca GP 400, La Bayette, France) were determined for each subject during a preliminary test session with the use of open-circuit spirometry (1100 medical gas analyzer, Marquette Electronics, Milwaukee, WI). Subsequently, all subjects performed a 90-min exercise at a workload corresponding to 68 ± 3% O_{2 max} (226 ± 3 vs. 140 ± 5 W for trained and sedentary subjects, respectively). One week later, trained subjects performed a 90-min exercise at a workload identical to that sustained by the sedentary subjects (140 ± 5 W corresponding to 47 ± 2.5% O_{2 max}). For this purpose, trained and sedentary subjects were matched by ranking them on the basis of O_{2 max} in their respective group. The exercises were performed between 9:30 and 11:00 AM in a laboratory with controlled temperature and humidity (21 ± 1°C, 45 ± 5%, respectively).

During each of the experiments, the subjects ingested 1,000 ml of room-temperature water containing 100 g of [¹³C]glucose. The drink was given in four equal volumes (25 g of glucose in 250 ml) taken 30 min before the beginning of exercise, immediately after the beginning of exercise, and at 30 and 60 min during the exercise period.

Assessment at rest. The subjects were also studied at rest 1 wk after the last exercise session. The subjects sat comfortably for 3.5 h in a reclining chair (from 8:30 to 12:00 AM). After basal measurements of O₂ and CO₂ production (CO₂) and blood sampling, the subjects ingested 100 g of [¹³C]glucose in a 10% solution (1,000 ml) divided in four equal volumes taken at 30-min intervals. Observations were made between 9:00 and 12:00 AM.

¹³C-labeling. Naturally labeled glucose (Biopharm, Laval, Canada) was artificially enriched with [U-¹³C]glucose (¹³C/C >99%; Isotec, Miamisburg, OH) to achieve a final isotopic composition larger than +25 and +55 change () in ¹³C Pee Dee Belemnitella-1 (PDB-1) at exercise and rest, respectively (actual values measured by mass spectrometry: +26.9 and + 57.5 ¹³C PDB-1). This high enrichment of exogenous glucose provided a very strong signal in expired CO₂ and allowed the neglect of the comparatively small changes in background ¹³C enrichment of expired CO₂ observed in response to exercise (31). The last evening meal before each experiment at rest or exercise (7:00 PM; 1,250 kcal: 70% carbohydrates, 15% fat, and 15% proteins) and the morning breakfast (7:30 AM; 500 kcal: 50% carbohydrates, 35% fat, and 15% proteins) were standardized. In addition, to keep a low background ¹³C enrichment of plasma glucose and expired CO₂, ingestion of carbohydrates from plants with the C₄ photosynthetic cycle, which are naturally enriched in ¹³C (19), was avoided 1 wk before and during the period of experiments. Subjects also refrained from exercising and from drinking coffee and alcohol.

Measurements and computations. Measurements and blood sampling were made every 30 min during the experiments conducted at rest and exercise (see Fig. 1). Glucose and fat oxidation were computed from indirect respiratory calorimetry corrected for protein oxidation. For this purpose, CO₂ and O₂ were measured by using open-circuit spirometry (10- and 5-min collection periods at rest and during exercise, respectively), and urea production was estimated over the observation periods. Urea concentration was measured in urine at rest and in urine and sweat during exercise. The amount of urine produced over the observation period (rest or exercise) was measured, and the volume of sweat produced during exercise was estimated from change in body mass, taking into account fluid and glucose intake, weight loss through CO₂, and water loss in the lungs (20). For the measurement of ¹³C/¹²C in expired CO₂, 80-ml samples of expired gases were collected and stored in vacutainers (Becton Dickinson, Franklin Lakes, NJ) until analysis. Blood samples (6 ml) were withdrawn at regular intervals through a catheter (Baxter Health Care, Valencia, CA) inserted into an antecubital vein at the beginning of the experiment for the measurement of plasma glucose, lactate, insulin, free fatty acid, and glycerol concentrations and for the measurement of ¹³C/¹²C in plasma glucose. Plasma, urine, and sweat samples were stored at 80°C until analysis.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Isotopic composition of expired CO2 and of plasma glucose and percentage of plasma glucose derived from exogenous glucose at rest (A and B) and during exercise (C and D), performed at the same relative and absolute workloads in trained and sedentary subjects. Exercise started at time 0. 13C PDB-1, difference in 13C Pee Dee Bilemnitella-1 (PDB-1) standard. Values are means ± SE; a significantly different from sedentary subjects; b significantly different from trained subjects at 68% maximal oxygen consumption (O2 max) (P  0.05).

Plasma glucose, lactate (Sigma Diagnostics, Mississauga, Canada), fatty acid, and glycerol (Boehringer Mannheim, Germany) concentrations were measured by using spectrophotometric automated assays, whereas plasma insulin concentration was measured by using an automated radioimmunoassay (KTSP-11001, Immunocorp Sciences, Montreal, Canada). Urea concentration in urine and sweat was measured by using a Synchron Clinical System (CX7, Beckman, Anaheim, CA).

For the measurement of ¹³C/¹²C in plasma glucose, 1 ml of plasma was first deproteinized with barium hydroxide (1.5 ml, 0.3 N) and zinc sulfate (1.5 ml, 0.3 N). The soluble phase was separated from the protein precipitate by centrifugation (20 min, 3,000 g, 4°C), and the remaining protein precipitate was washed with 3 ml of distilled water. The glucose was then separated by double-bed ion-exchange chromatography by running the combined supernatants (~7 ml) through superposed columns (0.5 × 2 cm) of AG 50W-X8 H⁺ (200-400 mesh) and (0.5 × 2 cm) of AG 1-X8 chloride (200-400 mesh) resins (Biorad, Mississauga, Canada) equilibrated and eluted with distilled water. The solution obtained (~10 ml) was evaporated to dryness (Virtis Research Equipment, New York, NY). The average percent recovery of glucose was 86 ± 3%. The eluate was then combusted for 60 min at 400°C in the presence of copper oxide (20 mg), and the CO₂ recovered was analyzed by mass spectrometry. This procedure was validated by Wolfe et al. (40) and yielded similar values of [¹³C]glucose enrichment than those obtained by gas chromatography-mass spectrometry or by the more specific isolation of glucose by crystallization as potassium gluconate. We have also observed that the material obtained after evaporation is not significantly contaminated by non-glucose carbons (29).

Measurements of ¹³C/¹²C in expired CO₂ and CO₂ from combustion of glucose extracted from the plasma were performed by mass spectrometry (Prism, VG, Manchester, UK) after cryodistillation, as previously described (1). The isotopic composition was expressed in difference by comparison with the PDB-1 Chicago standard where R_spl and R_std are the ¹³C/¹²C ratio in the sample and standard (1.12372%), respectively (7).

Protein oxidation was computed from the estimated amount of urea excreted during the exercise period (neglecting the small changes in plasma urea concentration), taking into account that 1 g of urea corresponds to 2.9 g of proteins oxidized (21). Glucose (G_total) and fatty acid (F_total) oxidation rates were then computed from O₂ and CO₂ (30) corrected for the volume of O₂ and CO₂ corresponding to protein oxidation (1.010 and 0.843 l/g, respectively)

(1)

(2)

with mass in grams and volume in liters (STPD). The contribution of the oxidation of the various substrates to the energy yield was computed from their respective energy potential at 37°C [3.87, 9.75, and 4.704 kcal/g for carbohydrate, fat, and proteins, respectively (21, 30)].

The oxidation rate of exogenous glucose (G_exo, in g/min) was computed as follows (31)

(3)

where CO₂ is in liters STPD per minute, R_exp is the ¹³C/¹²C observed in expired CO₂, R_ref is the ¹³C/¹²C in expired CO₂ at rest before exercise, R_exo is the ¹³C/¹²C in the artificially labeled exogenous glucose ingested, k₁ (0.7426 l/g) is the volume of CO₂ provided by the complete oxidation of glucose (31), and k₂ is the fractional recovery at the mouth of the CO₂ produced in tissues [0.8 and 1.0 at rest and exercise, respectively (3, 27)]. Because of the presence of a large bicarbonate pool in the body, the ¹³C/¹²C in expired CO₂ only slowly equilibrates with ¹³C/¹²C in the CO₂ produced in the tissues (27), particularly at rest, because of the small volume of CO₂ produced. To take into account this delay between ¹³CO₂ production in the tissues and at the mouth, the above computations were only made during the last hour of the observation periods, both at rest and during exercise.

Based on the isotopic composition of plasma glucose (R_glu), the percentage of plasma glucose derived from exogenous glucose (F_exo) and the oxidation rate of blood-borne glucose (G_blood) (11) were computed as follows

(4)

and

(5)

where R_glu-ref is the isotopic composition of plasma glucose observed before ingestion of labeled glucose. These values were not significantly different from each other and not significantly different from R_ref (see Fig. 1, A and B). The amount of glucose oxidized that was derived from glycogen in peripheral tissues (mainly muscle glycogen during exercise), either directly or through the lactate shuttle, was computed as the difference between the total amount of glucose oxidized (G_total, Eq. 1) and the amount of plasma glucose oxidized (G_blood, Eq. 5). Finally, the amount of glucose released from the liver that was oxidized was estimated as the difference between G_blood and G_exo.

Statistics. Data are presented as means ± SE. The main effects of time and training status as well as time-training status interactions were tested by repeated-measures analysis of variance (Statistica package). Newman-Keuls post hoc tests were used to identify the location of significant differences (P  0.05) when the analysis of variance yielded a significant F-ratio.

	RESULTS

Observations at rest. When labeled glucose was ingested at rest, the progressive increase in ¹³C enrichment of expired CO₂ was slightly but significantly faster in trained than in sedentary subjects, and higher values were reached during the last hour of the 180-min observation period (average values over the last hour: 1.0 ± 0.9 vs. 5.3 ± 1.5 ¹³C PDB-1) (Fig. 1A). The progressive enrichment of plasma glucose was similar in both groups. However, over the last hour of the observation period, the average value was slightly but significantly lower in trained vs. sedentary subjects (29.8 ± 1.8 vs. 34.6 ± 1.0 ¹³C PDB-1). Accordingly, the percentage of plasma glucose derived from ingested glucose was slightly but significantly lower in trained subjects (65 ± 2 vs. 71 ± 1% in sedentary subjects) (Fig. 1B).

Over the last hour of the 3-h observation period at rest, the amounts of proteins, fatty acids, and glucose oxidized were similar in trained and sedentary subjects and contributed 12, 24-34, and 64-54% to the energy yield, respectively (Fig. 2). In contrast, the amounts of exogenous glucose and of glucose derived from the liver (and, thus, the total amount of plasma glucose that was oxidized) were significantly higher in trained than in sedentary subjects.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Total energy substrate oxidation (A) and glucose oxidation (B) over the last hour of 180-min observation period after glucose ingestion at rest. Exo, exogenous; Periph, periphery. Values are means ± SE; a significantly different from sedentary subjects (P  0.05).

Changes in plasma glucose and insulin concentrations during the 180-min observation period are depicted in Fig. 3. The peak plasma glucose concentration was higher and was reached later in sedentary than in trained subjects, and the response of plasma insulin concentration was also markedly higher, although these differences did not reach statistical significance. However, the areas computed under the curve of plasma glucose and insulin concentrations were significantly larger in sedentary than in trained subjects (Fig. 3).

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Plasma glucose (A) and insulin (B) responses and area under curves (AU, arbitrary units) after glucose ingestion at rest in trained and sedentary subjects. Values are means ± SE; a significantly different from sedentary subjects (P  0.05).

Observations during exercise. For the same relative workload (68 ± 3 and 68 ± 4% O_{2 max} in trained and sedentary subjects, respectively), the mechanical power output and O₂ were significantly higher in trained than in sedentary subjects (226 ± 3 vs. 140 ± 5 W and 3.13 ± 0.12 vs. 2.35 ± 0.09 l/min, respectively). When the subjects were working at the same workload (140 ± 5 W), the O₂ was not significantly different in the two groups (2.20 ± 0.08 and 2.35 ± 0.09 l/min for trained and sedentary subjects, respectively).

Over the last hour of the 90-min exercise period, the amount of proteins oxidized was not significantly different in the three experimental conditions (6.0 ± 0.6 and 9.8 ± 1.3 g in trained subjects at 68 and 47% O_{2 max}; 6.9 ± 0.9 g in sedentary subjects at 68% O_{2 max}) (Fig. 4). The contribution of protein oxidation to the energy yield remained small and was not significantly different in the three experimental conditions (3-7%).

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Total energy substrate oxidation (A), glucose oxidation (B), and blood-borne glucose oxidation (C) over the last hour of 90-min exercise performed at same relative (68% O2 max) and absolute workloads (140 W: 47 vs. 68% O2 max for trained and sedentary subjects, respectively) in trained and sedentary subjects. Values are means ± SE; a significantly different from sedentary subjects; b significantly different from trained subjects at 47% O2 max (P  0.05).

For the same relative workload (68% O_{2 max}), the amounts of glucose and fatty acids oxidized were consistently higher in trained than in sedentary subjects (Fig. 4). However, glucose and fatty acid oxidation levels were similar in trained and sedentary subjects when expressed as percentage of the total energy yield (glucose: 56 ± 3 vs. 58 ± 2%; fatty acids: 41 ± 3 vs. 37 ± 2% for trained and sedentary subjects, respectively).

For the same absolute workload (140 ± 5 W or 47 and 68% O_{2 max} in trained and sedentary subjects, respectively), the amount of glucose oxidized was slightly lower in trained subjects, but the difference did not reach statistical significance (Fig. 4). The oxidation of fatty acids was similar in the two groups. The contributions of glucose (53 ± 2 vs. 58 ± 2%) and fatty acid oxidation (40 ± 2 vs. 37 ± 2%) to the energy yield were also similar in trained and sedentary subjects, respectively.

Compared with the values observed in sedentary subjects (33.6 ± 1.2 g), the amount of exogenous glucose oxidized over the last hour of exercise was significantly higher in trained subjects at the same relative (44.4 ± 1.8 g) and absolute workloads (39.0 ± 1.8 g) (Fig. 4). The contribution of exogenous glucose oxidation to the energy yield was similar in trained and sedentary subjects for the same relative workload (17.9 ± 0.6 and 18.0 ± 0.3%, respectively) and significantly higher in trained subjects for the same absolute workload (22.5 ± 1.2 vs. 18.0 ± 0.3%).

The percentage of plasma glucose arising from exogenous glucose, computed from ¹³C enrichment of plasma glucose, progressively rose from ~30% at 15 min before exercise to ~60-75% during the last 30 min of exercise (Fig. 1D). These values were significantly higher in trained than in sedentary subjects at rest before exercise (33 vs. 26% at 15 min before exercise) as well as at the same relative and absolute workloads (70-74 vs. 62% over the last 30 min of exercise). For the same relative workload, the amount of liver glucose oxidized over the last hour of exercise was similar in trained and in sedentary subjects (22.2 ± 3.0 vs. 24.0 ± 1.8 g), whereas it was slightly but not significantly lower in trained subjects for the same absolute workload (16.8 ± 2.4 vs. 24.0 ± 1.8 g, P = 0.1) (Fig. 4). For the same absolute workload, the amount of glucose arising from muscle glycogen that was oxidized over the last hour of exercise was significantly lower in trained (36.0 ± 3.0 g) than in sedentary subjects (51.0 ± 5.4 g) (Fig. 4). For the same relative workload, a much higher amount was oxidized in trained subjects (73.8 ± 7.2 g).

Changes in plasma glucose, lactate, glycerol, free fatty acid, and insulin concentrations are shown in Fig. 5. After a transient increase in response to glucose ingestion before exercise, plasma glucose and insulin concentrations fell at the onset of exercise and then stabilized around (insulin) or slightly above (glucose) basal values. No significant change in plasma lactate concentration was observed in response to exercise in trained subjects both at 68 and 47% O_{2 max}, whereas a significant increase was observed in sedentary subjects. Plasma glycerol concentration increased over basal values in response to exercise at 68% O_{2 max} in both groups. In contrast, plasma free fatty acid concentration only increased in sedentary subjects exercising at 68% O_{2 max}.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Plasma glucose (A), insulin (B), lactate (C), glycerol (D), and free fatty acid (E) concentrations at rest and during exercise performed at same relative and absolute workloads in trained and sedentary subjects: exercise started at time 0. Values are means ± SE; a significantly different from sedentary subjects; b significantly different from trained subjects at 47% O2 max (P  0.05).

	DISCUSSION

Results from the present experiment indicate that, compared with sedentary subjects, exogenous glucose was oxidized at a higher rate both at rest and at exercise performed at the same relative and absolute workloads in trained subjects. After ingestion of 100 g of ¹³C-labeled glucose, compared with sedentary subjects, the amount of exogenous glucose oxidized over the last hour of the observation period was 25% higher at rest (8.4 ± 0.3 vs. 6.6 ± 0.8 g), 32% higher for the same relative workload (68% O_{2 max}: 44.4 ± 1.8 vs. 33.6 ± 1.2 g), and 16% higher for the same absolute workload (140 W or 47 vs. 68% O_{2 max}: 39.0 ± 1.8 vs. 33.6 ± 1.2 g) in trained subjects.

In the present experiment, compared with sedentary subjects, the response of plasma glucose and insulin concentrations after glucose ingestion at rest were lower in trained subjects. This phenomenon has already been shown, for example, by Lohman et al. (22) and Seals et al. (38) during an oral glucose tolerance test. Results from studies using euglycemic-hyperinsulinemic clamp techniques indicate that this is due to an increased sensitivity and responsiveness to insulin in trained vs. untrained subjects (25, 36) or after training (9). Data from Dela et al. (9) obtained at the first stage of a three-step hyperinsulinemic-euglycemic clamp (plasma insulin concentration = 387 pmol/l, which is comparable to the values observed in the present experiment after glucose ingestion at rest) indicate that the increased sensitivity to insulin in a trained leg was associated with a higher glycogen storage and lactate release and with an increased carbohydrate oxidation compared with the untrained contralateral leg.

The metabolic fate of exogenous glucose administered at rest has been described by several authors. When 100 g of glucose are ingested, the contribution of exogenous glucose to the energy yield ranges between 19 and 27% in nonobese subjects (15, 26, 27, 34), with values significantly lower in obese subjects with a poor glucose tolerance (14%) (34). The effect of training on the oxidative fate of an oral glucose load at rest has only been described by Krzentowski et al. (18). In this longitudinal study, subjects underwent an oral glucose tolerance test (100 g of naturally labeled glucose) before and after 6 wk of moderate exercise training (1 h of cycling, 5 days/wk, 30-40% O_{2 max}). The amount of exogenous glucose oxidized during the 7-h observation period was not significantly modified (35.9 ± 2.1 vs. 37.4 ± 2.0 g/7 h before and after training, respectively), contributing 27% to the energy yield.

In the present experiment, the average contribution of exogenous glucose oxidation to the energy yield at rest was 22.4 ± 2.2% in sedentary subjects, but 28.1 ± 1.1% in trained subjects, with a better glucose tolerance and both lower plasma glucose and insulin responses to the glucose load. This better disposal of exogenous glucose was associated with a higher oxidation rate of both plasma (~35%) and exogenous glucose (~25%). These findings are in line with data from Dela et al. (9) observed at the first stage of a hyperinsulinemic-euglycemic clamp, indicating that glucose uptake and oxidation across the trained leg were ~35% and ~30% higher, respectively. In the study by Dela et al., it was not possible, however, to distinguish between the sources of glucose oxidized: blood-borne glucose vs. muscle glycogen. Results from the present study indicate that the oxidation of blood-borne glucose is actually favored in trained subjects. This phenomenon could in part contribute to the better disposal of an oral glucose load.

The higher oxidation rate of exogenous glucose observed in the present experiment was associated with a slightly lower enrichment of plasma glucose in trained subjects. This could indicate that glucose released from the liver was reduced in a lesser extent after glucose ingestion. No data are presently available to support this suggestion on the effect of training on plasma glucose kinetics at rest during an oral glucose challenge. In the present experiment, the oxidation rate of liver glucose was much higher in trained than in sedentary subjects. However, liver glucose oxidation could not parallel liver glucose output because of possible changes in nonoxidative glucose disposal.

In the longitudinal study by Krzentowski et al. (17), the oxidation of an oral load of 100 g of [¹³C]glucose was measured over a 90-min exercise bout performed at 40% of the pretraining O_{2 max} (1.3 l/min), before and after training. As estimated from their Fig. 4 reported, the average oxidation rate of exogenous glucose over the exercise period was ~10% higher after than before training (~0.33 vs. 0.29 g/min), with exogenous glucose contributing ~18 and ~22% to the energy yield, before and after training, respectively. Jeukendrup et al. (16) have recently measured the oxidation of glucose ingested during exercise in trained and sedentary subjects. In this study, subjects performed a 120-min bout of cycling at 50% of the maximal mechanical power output (~50 and ~63% O_{2 max} in trained and sedentary subjects, respectively) while ingesting 100 g of glucose naturally enriched in ¹³C. During the last hour of exercise, although the difference did not reach statistical significance, the oxidation rate of exogenous glucose was also ~10% higher in trained than in sedentary subjects.

Results from the present experiment, observed at similar absolute and relative workloads in trained and sedentary subjects, are in line with the early findings by Krzentowski et al. (18) and the more recent findings by Jeukendrup et al. (16). Over the last hour of the 90-min exercise, the average rate of exogenous glucose was higher in trained subjects for the same relative (0.74 ± 0.04 vs. 0.56 ± 0.02 g/min in sedentary subjects) and absolute workloads (0.65 ± 0.04 vs. 0.56 ± 0.02 g/min in sedentary subjects). Previous work from our laboratory (23) as well as from Pirnay et al. (33) and an analysis of data in literature (29) indicate that the oxidation rate of exogenous glucose during exercise increases with workload and energy expenditure and, for a 100-g load, averages ~0.45 g/min and 0.65 g/min for O₂ = 2.2 and 3.1 l/min, respectively, contributing ~17% to the energy yield (29). In the present experiment, for a given relative workload (68% O_{2 max}), the higher oxidation rate of exogenous glucose observed in trained subjects can thus be explained by a 60% higher workload (226 ± 3 vs. 140 ± 5 W and 3.13 ± 0.12 vs. 2.35 ± 0.09 l O₂/min). In fact, the contribution of exogenous glucose oxidation to the energy yield was similar in both groups (17.9 ± 0.7 vs. 18.0 ± 0.3% for trained and sedentary subjects, respectively). In contrast, for the same absolute workload (140 ± 5 W), despite similar O₂ (2.20 ± 0.08 vs. 2.35 ± 0.09 l/min in trained and sedentary subjects, respectively), the average oxidation rate of exogenous glucose was higher in trained subjects. As a consequence, a larger percentage of the energy yield was derived from exogenous glucose oxidation (22.5 ± 1.2 vs. 18.0 ± 0.3%).

Dela et al. (10) showed that training increases the capacity of skeletal muscle to take up plasma glucose under the combined stimulation of a large dose of insulin (plasma insulin concentration: 18,500 pmol/l) and exercise. However, when plasma insulin remains in a physiological range, glucose uptake during exercise appeared similar in a trained and untrained leg (12, 37). Furthermore, consistent observations (4, 5, 24, 32) show that whole body glucose turnover and utilization for a given absolute workload are actually lower in trained than in sedentary subjects, thus resulting in a significant sparing of liver glycogen stores (5). Results from the present experiment are in line with these findings and indicate that glucose from the liver was oxidized to a lesser extent in trained than in sedentary subjects, even when comparatively large amounts of glucose were ingested during exercise. Indeed, the higher oxidation rate of exogenous glucose observed for a given absolute workload in trained vs. sedentary subjects was not associated with a higher oxidation rate of plasma glucose, which was similar in both groups (0.97 ± 0.03 vs. 0.95 ± 0.08 g/min in trained and sedentary subjects, respectively). Accordingly, the average oxidation rate of glucose released from the liver was 43% lower in trained subjects for a given absolute workload (although this did not reach statistical significance; P = 0.1). As expected (6), the oxidation rate of muscle glycogen was significantly (41%) lower. For the same relative workload, despite a much higher energy expenditure and a much higher rate of fatty acid (+45%) and muscle glycogen oxidation (+44%), the oxidation rate of glucose from the liver was similar in both groups (0.37 ± 0.05 vs. 0.40 ± 0.03 g/min in trained and sedentary subjects, respectively).

Taken together, these data indicate that when trained subjects are fed comparatively large amounts of glucose during exercise, the oxidation of exogenous glucose is selectively favored. This phenomenon results in a lower oxidation rate of glucose from the liver and could also contribute to the lower oxidation rate of glucose from muscle glycogen.