Journal Logo

Applied Sciences: Physical Fitness and Performance

Effect of timing of carbohydrate ingestion on endurance exercise performance


Author Information
Medicine & Science in Sports & Exercise: October 1996 - Volume 28 - Issue 10 - p 1300-1304
  • Free


The ingestion of carbohydrate during prolonged strenuous (70-75%˙VO2 peak) exercise has been shown to enhance endurance exercise performance (2,9,10). It appears that this improvement is due to maintenance of blood glucose levels and carbohydrate oxidation late in exercise when muscle glycogen levels are low(6,9). Carbohydrate ingestion increases glucose uptake during prolonged strenuous exercise but has little influence on net muscle glycogen utilization(4,9,15,19). The elevation in glucose uptake with carbohydrate ingestion widens as exercise duration increases(19). Indeed, the infusion rate necessary to maintain hyperglycemia (11,17) and euglycemia(17) during prolonged exercise increases with exercise duration. A high proportion of glucose uptake is oxidized by the contracting musculature (8,17). It appears possible, therefore, that carbohydrate ingestion may be of most benefit late in exercise when muscle glycogen levels are greatly reduced and the muscle's capacity to take up glucose is high.

Although it is generally accepted that carbohydrate ingestion increases exercise time to exhaustion at 70-75% ˙VO2max(2,9,10,13), the effects of carbohydrate ingestion on performance following prolonged exercise are less clear. Several studies have used an exercise performance protocol in which subjects exercised for 105-120 min at approximately 70% ˙VO2max with and without carbohydrate ingestion and then performed as much work as possible in 8-15 min. Such an exercise model is likely to be closer to what might be experienced in competition compared with exercise time to fatigue. Carbohydrate ingestion has been found to improve performance in some(20,21) but not all (23) studies employing this type of protocol. This study examined the efficacy of carbohydrate ingestion throughout exercise compared with late in exercise, on 15-min work output following 2 h of exercise at 70% ˙VO2 peak. Ingestion of a large bolus of carbohydrate (200 g in a 50% solution) within 30 min of the point of fatigue during prolonged exercise at 70%˙VO2max has been shown to increase exercise time to exhaustion to a extent similar to that with carbohydrate ingestion throughout exercise(7). In this study we hypothesized that ingestion of a large amount of carbohydrate late in exercise would improve 15-min performance as much or greater than when carbohydrate is ingested throughout exercise.


Subjects. Eight well-trained male cyclists/triathletes (23± 1 yr, 71.7 ± 1.4 kg, mean ± SEM) volunteered for this study, which had been approved by The University of Melbourne Human Research Ethics Committee. Before the study began, the subjects completed a medical questionnaire and provided written informed consent. Approximately 1 wk prior to the first trial, peak pulmonary oxygen uptake (˙VO2 peak) was determined during incremental cycling (Lode, Groningen, The Netherlands) to volitional fatigue and averaged 4.94 ± 0.14 l·min-1. The subjects were then given a 30-min recovery before undertaking a 15-min“all-out” ride to familiarize them with the performance component of the experimental trials. Diet and exercise were controlled over the 24 h prior to each trial. Approximately 24 h prior to each trial, subjects reported to the laboratory for a 45-min cycling bout at 70 ± 1% ˙VO2 peak. They then refrained from strenuous exercise and were supplied with food for the remainder of the day (14.0 MJ: 82% CHO, 6% fat, 12% protein). In addition, the subjects were instructed to refrain from caffeine, alcohol, or tobacco consumption during the 24 h prior to a trial. We have previously found that such pre-exercise diet and exercise control results in reproducible metabolite and hormone levels.

Experimental procedures. The subjects reported to the laboratory in the morning after an overnight fast or in the afternoon at least 6 h post-prandial. A catheter was inserted into an antecubital vein for blood sampling. After 20 min of seated rest, a pre-exercise blood sample was obtained. The subjects then cycled for 5 min at 150 W before the ergometer workload was increased to a workload requiring 70 ± 1% ˙VO2 peak. They then exercised at this workload for 2 h followed immediately(without stopping exercise) by a 15-min “all-out” performance ride. The performance ride was executed with the ergometer placed in the linear mode (cadence vs workload) with the slope of the linear function tailored to each individual based on his trial workload and preferred cycling cadence. Since the level of investigator encouragement can affect subject performance, care was taken that each subject was encouraged to a similar extent by the same investigator in each trial. The catheter for blood sampling was kept patent by flushing with 0.9% saline and every 30 min with 0.5 ml of saline containing 5 units of heparin. The laboratory was maintained at 20-22°C and a fan circulated air to minimize thermal stress. In three trials subjects ingested 250 ml of fluid immediately upon beginning exercise and then 250 ml every 15 min of exercise including the 120-min time point. On one occasion (CHO-7) subjects ingested a 7% commercial carbohydrate-electrolyte drink (Sport Plus, Cadbury Schweppes Pty. Ltd., Melbourne, Australia) throughout exercise, while in a second trial (CON) an artificially sweetened placebo was ingested throughout exercise. In a third trial (CHO-0/21), the placebo was consumed at 0, 15, 30, 45, 60, and 75 min, then a 21% glucose polymer drink was given at 90, 105, and 120 min. The 21% carbohydrate drink was composed of the placebo with 21 g·100 ml-1 of a glucose polymer added. The amount of carbohydrate (157.5 g) ingested was the same in the CHO-7 and CHO-0/21 trials. The three trials were randomly assigned and conducted double blind. The trials were separated by a minimum of 5 d. Every 30 min during exercise, heart rate (Accurex, Polar, Finland) was monitored and expired air collected into Douglas bags for oxygen uptake and respiratory exchange ration (RER) determination. Carbohydrate oxidation was estimated from oxygen consumption and RER using standard formulae. Heart rate and expired gas samples were also obtained at 5-min intervals during the 15-min performance ride. The subjects were unable to see their heart rates at any stage during the trials, and they were not informed of any aspect of their performances until after the conclusion of the three trials. Blood samples were obtained at 0, 30, 60, 90, 105, 120, and 135 min of exercise to measure plasma glucose. In addition, samples were obtained at 0, 15, and 120 min of exercise for measurement of hemoglobin to estimate blood volume changes during exercise (14). Plasma was also portioned for lactate analysis at 0, 60, 120, and 135 min of exercise and for insulin at 0, 60, and 120 min. The samples for glucose, insulin, and lactate were placed in fluoride heparin tubes. The tubes were spun and the plasma stored at -20°C for later analysis. For plasma lactate determination, 500μl of the plasma was deproteinized in 8% perchloric acid, spun, and stored at -20°C.

Analytical techniques. Expired air samples were measured for oxygen and carbon dioxide using Applied Electrochemistry S-3A/II and CD-3A electronic analyzers (Ametek, Pittsburgh, PA). These analyzers were calibrated using commercial gases of known composition. Volume was measured on a Parkinson-Cowan gas meter calibrated against a Tissot spirometer. Hemoglobin was measured in triplicate using the cyanmethemoglobin method, with changes in blood volume estimated using the Dill and Costill equation(14). Plasma glucose was measured using an automated glucose oxidase method (YSI 2300, Yellow Springs, OH). Plasma lactate was determined using a standard enzymatic technique (18), while plasma insulin was measured by radioimmunoassay (Incstar, Stillwater, MN). Data from the three trials were compared using two factor analysis of variance for repeated measures with significance at the P < 0.05 level. Specific differences were located using the Student-Newman-Keulspost-hoc test. All data are reported as means ± SEM.


Oxygen uptake was similar during the 2-h exercise bout at 70%˙VO2 peak in the three trials averaging 3.47 ± 0.04, 3.46± 0.04, and 3.46 ± 0.04 l·min-1 in CON, CHO-7, and CHO-0/21, respectively. Heart rate in the three trials was similar, averaging 150 ± 2, 148 ± 2, and 150 ± 2, in CON, CHO-7, and CHO-0/21, respectively. RER was higher (P < 0.05) over the second hour of exercise in CHO-7 compared with CON (0.94 ± 0.01 vs 0.92± 0.01 at 60 min, 0.93 ± 0.01 vs 0.90 ± 0.01 at 90 min, and 0.92 ± 0.01 vs 0.89 ± 0.01 at 120 min). In addition, RER was higher in CHO-0/21 compared with CON at 90 min (0.92 ± 0.01 vs 0.90± 0.01) and 120 min (0.91 ± 0.01 vs 0.89 ± 0.01). No differences in RER during the 2-h exercise bout were observed between CHO-7 and CHO-0/21. Plasma glucose levels were higher until 90 min in CHO-7 compared with CHO-0/21 and higher throughout exercise in CHO-7 compared with CON(Fig. 1). Plasma glucose levels were similar in CON and CHO-0/21 until 90 min. Ingestion of carbohydrate in CHO-0/21 resulted in plasma glucose levels rising above both CON and CHO-7 at 120 min and 135 min(Fig. 1). Plasma insulin levels were higher after 60 min of exercise in CHO-7 compared with CON and CHO-0/21 (Table 1). At 120 min of exercise, plasma insulin levels were similar in CHO-7 and CHO-0/21 with the levels in both being higher than those in CON(Table 1). Plasma lactate levels were similar in the three trials except at 60 min where they were higher in CHO-7 compared with CHO-0/21 (Table 1). Blood volume decreased to a similar extent in all trials during the first 15 min of exercise (-6.8 ± 1.1%,-6.7 ± 1.1%, and -6.8 ± 0.6% in CON, CHO-7, and CHO-0/21, respectively), then remained relatively stable, and not significantly different, for the remainder of the exercise bout in the three trials. At 120 min of exercise blood volume change was -7.8 ± 1.5%, -7.7 ± 1.1%, and -6.7 ± 1.3% in CON, CHO-7, and CHO-0/21, respectively. Data from the 15-min performance ride are shown in Table 2. Total work output was higher when carbohydrate was ingested throughout exercise (CHO-7) compared with CON. Seven of the eight subjects performed best in the CHO-7 trial. There was no difference in work output between CON and CHO-0/21 or between CHO-7 and CHO-0/21. The increased work output in CHO-7 was associated with a higher oxygen uptake, RER, and estimated carbohydrate oxidation during exercise compared with CON. In addition, carbohydrate oxidation and RER were higher in CHO-0/21 than CON while carbohydrate oxidation and RER were similar between CHO-7 and CHO-0/21. Plasma lactate levels at the end of the performance ride were not significantly different among the three trials (Table 1).


In contrast to our hypothesis, carbohydrate ingestion throughout exercise improved 15-min work output following prolonged exercise, but ingestion of an equal amount of carbohydrate late in exercise did not improve performance compared with the control trial (Table 2). Seven of the eight subjects performed best when they ingested carbohydrate throughout exercise. This result is somewhat surprising since blood glucose levels were actually higher during the performance ride in CHO-0/21 compared with CHO-7(Fig. 1), and muscle glycogen levels would be expected to be similar at the beginning of the performance ride in the three trials(9,15). In addition, plasma insulin levels were similar in CHO-7 and CHO-0/21 immediately prior to the performance ride(Table 1). Therefore, it is likely that carbohydrate availability was greater in CHO-0/21 than both CHO-7 and CON during the performance ride. Despite this, work output in CHO-0/21 was not greater than that achieved in CON. This suggests that effects of not consuming carbohydrate during the first 90 min of exercise carried over to the performance trial, despite an elevation in plasma glucose and insulin levels late in exercise. Ingestion of a large bolus of carbohydrate approximately 30 min prior to fatigue during strenuous exercise has been found to result in a 21% extension in time to exhaustion compared with ingestion of a placebo(7). Therefore, it appears the ability to maintain a constant work level is enhanced by carbohydrate ingestion late in exercise, while the ability to increase work output is not improved significantly. Only when carbohydrate was ingested throughout exercise was average˙VO2, and therefore exercise intensity, significantly increased during the 15-min performance ride (Table 2). Seven of eight subjects performed better in CHO-7 compared with CON, but only three performed better in CHO-0/21 compared with CON.

Others have also found that carbohydrate ingestion throughout exercise improves performance following 105-120 min of exercise at 70%˙VO2max(20,21). It has been suggested that this improvement is due to a maintenance of blood glucose levels and carbohydrate oxidation late in exercise when muscle glycogen levels are low(6,9). Indeed, RER was higher late in exercise in CHO-7 compared with CON and, therefore, the increased carbohydrate availability in CHO-7 compared with CON may have contributed to the higher carbohydrate oxidation rate in the 15-min performance ride, resulting in enhanced performance (Fig. 1; Table 2). However, it is difficult to explain why work output was not increased significantly in CHO-0/21 since blood glucose levels in that trial were higher at the end of exercise than those in both other trials and plasma insulin levels were similar to those in the CHO-7 trial. In addition, rates of carbohydrate oxidation during the latter stages of the 2-h submaximal work bout were similar in CHO-0/21 and CHO-7.

It appears likely that factors other than the level of carbohydrate availability influenced work output in CHO-0/21. There may have been a delay in the stimulative effect of the hyperglycemia and hyperinsulinemia on muscle glucose uptake in CHO-0/21, such that glucose uptake was not sufficient to improve exercise performance. Carbohydrate ingestion has been shown to stimulate glucose uptake with the effect increasing with increases in exercise duration (19). However, it has been shown that glucose infusion begun at the start of exercise results in hyperglycemia, hyperinsulinemia, and increased carbohydrate oxidation within 2 min of the onset of exercise (17). It is also possible that plasma FFA levels may have been elevated at 90 min in CHO-0/21 compared with CHO-7, resulting in an inhibition of glucose uptake during the performance ride(16).

Other factors such as alterations in the muscle membrane potential(22), impaired calcium homeostasis(5), and free radical induced muscle damage(1,12) have been implicated in fatigue during prolonged exercise. However, it is unlikely these processes would be greatly affected by carbohydrate ingestion. There is also little reason to suspect a difference in the rate of dehydration or hyperthermia since the rate of fluid consumption and blood volume changes were identical in the trials. Although speculative, it has been suggested that carbohydrate ingestion throughout exercise may attenuate the onset of negative central nervous system changes late in exercise (3). Carbohydrate ingestion results in an attenuation of the normal exercise-induced rise in plasma FFA, free-tryptophan, and the free-tryptophan to branched-chain amino acid ratio(13). These alterations with carbohydrate ingestion may result in a reduced degree of central fatigue owing to lower brain serotonin levels.

In conclusion, ingestion of carbohydrate throughout prolonged exercise improves the ability to produce work during a subsequent 15-min performance ride, while ingestion of the same quantity of carbohydrate late in exercise does not significantly improve work output compared with ingestion of a placebo. The inability to significantly improve work output when carbohydrate was ingested late in exercise occurred despite elevated plasma glucose and insulin levels. These results suggest that carbohydrate ingestion may improve performance through mechanisms other than, or in addition to, increased carbohydrate availability to contracting skeletal muscle.

Figure 1-Plasma glucose before and during 2 h of exercise at 70± 1% ˙VO2 peak and after a 15-min performance ride with ingestion of a sweet placebo throughout exercise (CON), 7% carbohydrate throughout exercise (CHO-7) or a sweet placebo for the first 90 min followed by 21% carbohydrate (CHO-0/21). Values are means ± SEM(
Figure 1-Plasma glucose before and during 2 h of exercise at 70± 1% ˙VO2 peak and after a 15-min performance ride with ingestion of a sweet placebo throughout exercise (CON), 7% carbohydrate throughout exercise (CHO-7) or a sweet placebo for the first 90 min followed by 21% carbohydrate (CHO-0/21). Values are means ± SEM(:
N = 8). * denotes different from CON( P < 0.05); # denotes different from CHO-0/21( P < 0.05).


1. Barclay, J. K. and M. Hansel. Free radicals may contribute to oxidative skeletal muscle fatigue. Can. J. Physiol. Pharmacol. 69:279-284, 1991.
2. Björkman, O., K. Sahlin, L. Hagenfeldt, and J. Wahren. Influence of glucose and fructose ingestion on the capacity for longterm exercise in well-trained men. Clin. Physiol. 4:483-494, 1984.
3. Blomstrand, E., F. Elsing, and E. A. Newsholme. Changes in plasma concentrations of aromatic and branched-chain amino acids during sustained exercise in man and their possible role in fatigue. Acta Physiol. Scand. 133:115-121, 1988.
4. Bosch, A. N., S. C. Dennis, and T. D. Noakes. Influence of carbohydrate ingestion on fuel substrate turnover and oxidation during prolonged exercise. J. Appl. Physiol. 76:2364-2372, 1994.
5. Byrd, S. K., A. K. Bode, and G. A. Klug. Effects of exercise of varying duration on sarcoplasmic reticulum function. J. Appl. Physiol. 66:1383-1389, 1989.
6. Coggan, A. R. and E. F. Coyle. Reversal of fatigue during prolonged exercise by carbohydrate infusion or ingestion. J. Appl. Physiol. 63:2388-2395, 1987.
7. Coggan, A. R. and E. F. Coyle. Metabolism and performance following carbohydrate ingestion late in exercise. Med. Sci. Sports Exerc. 21:59-65, 1989.
8. Coggan, A. R., R. J. Spina, W. M. Kohrt, D. M. Bier, and J. O. Holloszy. Plasma glucose kinetics in a well-trained cyclist fed glucose throughout exercise. Int. J. Sports Nutr. 1:279-288, 1991.
9. Coyle, E. F., A. R. Coggan, M. K. Hemmert, and J. L. Ivy. Muscle glycogen utilization during prolonged strenuous exercise when fed carbohydrate. J. Appl. Physiol. 61:165-172, 1986.
10. Coyle, E. F., J. M. Hagberg, B. F. Hurley, W. H. Martin, A. A. Ehsani, and J. O. Holloszy. Carbohydrate feeding during prolonged strenuous exercise can delay fatigue. J. Appl. Physiol. 55:230-235, 1983.
11. Coyle, E. F., M. T. Hamilton, J. Gonzalez-Alonso, S. J. Montain, and J. L. Ivy. Carbohydrate metabolism during intense exercise when hyperglycemic. J. Appl. Physiol. 70:834-840, 1991.
12. Davies, K. J. A., A. T. Quintanilha, G. A. Brooks, and L. Packer. Free radicals and tissue damage produced by exercise.Biochem. Biophys. Res. Comm. 107:1198-1205, 1982.
13. Davis, J. M., S. P. Bailey, J. A. Woods, F. J. Galiano, M. T. Hamilton, and W. P. Bartoli. Effects of carbohydrate feedings on plasma free tryptophan and branched-chain amino acids during prolonged cycling.Eur. J. Appl. Physiol. 65:513-519, 1992.
14. Dill, D. B. and D. L. Costill. Calculation of percentage changes in volumes of blood, plasma, and red cells in dehydration.J. Appl. Physiol. 37:247-248, 1974.
15. Hargreaves, M. and C. A. Briggs. Effect of carbohydrate ingestion on exercise metabolism. J. Appl. Physiol. 65:1553-1555, 1988.
16. Hargreaves, M., B. Kiens, and E. A. Richter. Effect of increased plasma free fatty acid concentrations on muscle metabolism in exercising men. J. Appl. Physiol. 70:194-201, 1991.
17. Hawley, J. A., A. N. Bosch, S. M. Weltan, S. C. Dennis, and T. D. Noakes. Glucose kinetics during prolonged exercise in euglycaemic and hyperglycaemic subjects. Pflügers Arch. 426:378-386, 1994.
18. Lowry, O. H. and J. V. Passonneau. A Flexible System of Enzymatic Analysis. New York: Academic Press, 1972, pp. 199-201.
19. McConell, G., S. Fabris, J. Proietto, and M. Hargreaves. Effect of carbohydrate ingestion on glucose kinetics during exercise. J. Appl. Physiol. 77:1537-1541, 1994.
20. Mitchell, J. B., D. L. Costill, J. A. Houmard, W. J. Fink, D. D. Pascoe, and D. R. Pearson. Influence of carbohydrate dosage on exercise performance and glycogen metabolism. J. Appl. Physiol. 67:1843-1849, 1989.
21. Murray, R., G. L. Paul, J. G. Seifert, and D. E. Eddy. Responses to varying rates of carbohydrate ingestion during exercise.Med. Sci. Sports Exerc. 23:713-718, 1991.
22. Sahlin, K. and S. Broberg. Release of K+ from muscle during prolonged dynamic exercise. Acta Physiol. Scand. 136:293-294, 1989.
23. Zachwieja, J. J., D. L. Costill, G. C. Beard, R. A. Robergs, D. D. Pascoe, and D. E. Anderson. The effects of a carbonated carbohydrate drink on gastric emptying, gastrointestinal distress, and exercise performance. Int. J. Sports Nutr. 2:239-250, 1992.


©1996The American College of Sports Medicine