Journal Logo

APPLIED SCIENCES: Psychobiology and Behavioral Strategies

Effect of carbohydrate ingestion on ratings of perceived exertion during a marathon


Author Information
Medicine & Science in Sports & Exercise: November 2002 - Volume 34 - Issue 11 - p 1779-1784
  • Free


Numerous studies have examined the effect of bodily carbohydrate substrate availability on the perception of exertion during prolonged strenuous exercise (3–5,8,11,12,17,20,21). Despite some exceptions (8,11), a majority of these studies have demonstrated that the perceived exertion is associated with a decrease in blood glucose and carbohydrate oxidation (3–5,12,17,20,21). The physiological link between perceived exertion and energy substrate may be mediated through intensified sensory input arising from skeletal muscle owing to a depletion of carbohydrate energy sources (15).

The evidence supporting the relation between the strength of perceived exertion and availability of carbohydrate energy substrate was derived from well-controlled laboratory studies in which subjects were given carbohydrate or placebo while exercising at a given level of intensity until exhaustion. Whether the availability of carbohydrate energy substrate will also affect the strength of perceived exertion during a sports competition, such as a marathon race, in which intensity of exercise is self-selected, has not yet been documented. This is an intriguing question as running is a mode of exercise for which the attenuating effect of carbohydrate energy substrate on perceived exertion remains controversial (20,21). In addition, a number of studies have failed to observe an ergogenic effect of carbohydrate supplementation on marathon running (14,18), suggesting that factors other than availability of carbohydrate energy substrates may be comparatively more important in mediating perceived exertion during running of varying intensities.

The present study was undertaken to examine whether and/or how the availability of carbohydrate energy substrate affects the perception of exertion during a competitive marathon race. The internal validity of this field-based study was ensured by using a randomized and double-blind experimental design. This research approach was similar to those of our prior investigations (20,21), except that in the present study the intensity of exercise was self-selected by the runners and was conducted in a field setting. It was our hypothesis that carbohydrate ingestion would attenuate ratings of perceived exertion in the presence of enhanced running performance.



Marathon runners were recruited through a letter of invitation before the April 10, 1999, Charlotte Marathon in Charlotte, NC, and the July 8, 2000, Grandfather Mountain Marathon in Boone, NC. Male and female runners ranging in age from 21 to 72 yr were accepted into the study if they had run at least one competitive marathon and were willing to adhere to all aspects of the research design including randomization to either the carbohydrate or placebo group. Informed written consent was obtained from each subject, and the experimental procedures were in accordance with the policy statements of the institutional review board of Appalachian State University (ASU) and the American College of Sports Medicine.

Research design.

The same research design and procedures were used for both marathon race events, and the data were combined. Two to four weeks before the marathon race events, subjects reported to the ASU Human Performance Lab for orientation and measurement of height, weight, body composition, and maximal O2 uptake (V̇O2max). Body composition was assessed from hydrostatic weighing and V̇O2max determined using a graded maximal treadmill adapted for runners as described in earlier studies from our group (20,21). Oxygen uptake and ventilation were measured using a MedGraphics CPX metabolic system (MedGraphics Corporation, St. Paul, MN). Analyzers were calibrated using gases provided by MedGraphics Corporation: 1) calibration gas: 5%CO2, 12%O2, balance N2; and 2) reference gas 21%O2, balance N2. The standard specification of error for the reference and calibration gas is ± 0.10% (MedGraphics Corporation). Maximal heart rate was measured using a chest heart rate monitor (Polar Electro Inc., Woodbury, NY). Training history and demographic factors were assessed through a questionnaire. During orientation, a dietitian instructed the runners to follow a high-carbohydrate diet, record intake in a food record during the 3 d before the race events, and avoid food or beverages containing calories or caffeine from 9:00 p.m. the night preceding the race. Nutrient intake was assessed using the computerized dietary analysis system, Food Processor Plus, version 6.0 (ESHA Research, Salem, OR).

During maximal testing, ratings of perceived exertion (RPE) were determined using the Borg 15-point rating of perceived exertion scale to establish the low and high rating anchors (1). The definition of RPE and instructions for the use of the rating scale were read by the subject before the exercise test. The instructions emphasized that the perceptual ratings should reflect sensations of exertion, strain, discomfort, and/or fatigue in the limbs and respiratory system. Each rating was limited to a single numbered response. Each subject also received a pocket-sized RPE scale to be used during training in the 2–4 wk before the race. The distribution of the pocket-sized RPE scale was done to promote familiarization with the rating procedure during training and to enhance recall capabilities of the perceptual responses when assessed during the actual marathon.

On the race days, 102 subjects reported to the start area in a 9-h fasted state at 5:00–6:00 a.m. After sitting for 10–15 min, blood samples were collected (i.e., for plasma volume, glucose, lactate, cortisol, insulin, growth hormone, epinephrine, and norepinephrine). Body mass was measured, and a chest heart rate monitor was attached to each runner. The runners were randomly assigned to carbohydrate or placebo groups, with beverage plastic bottles containing the experimental beverages administered in double-blind fashion using color codes. The beverages were supplied by the Gatorade Sports Science Institute (Barrington, IL). The 6% carbohydrate and placebo beverages were identical in appearance and taste. The two fluids were identical in sodium (∼19.0 mEq·L−1) and potassium (∼3.0 Eq·L−1) concentration, and pH (∼3.0). Each runner ingested 650 mL of beverage approximately 30 min before the start of the races (7:00 a.m.). During the race, runners drank approximately 1000 mL of beverage each hour. Research assistants were positioned every 3.2 km along the racecourse to deliver the placebo/experimental beverage bottles, which contained 500 mL of fluid. Runners ingested the fluid from two bottles per hour. Runners agreed to avoid all other beverages and food before and during the race. The research assistants also recorded heart rates and RPE from each runner every 3.2 km. After runners crossed the race finish line, blood samples were collected from each subject within 5 min. Body mass was also measured postrace.

Hormones, glucose, lactate, and plasma volume.

Blood samples were drawn from an antecubital vein with subjects in the seated position. Plasma cortisol was assayed using a competitive solid-phase 125I radioimmunoassay (RIA) technique (Diagnostic Products Corporation, Los Angeles CA). RIA kits were also used to determine plasma concentrations of insulin and growth hormone according to manufacturer’s instructions (Diagnostic Products Corporation). Plasma was analyzed spectrophotometrically for glucose (prerun and immediate postrun samples) (10). Lactate was measured from finger-stick blood samples by using a lactate analyzer (YSI 2300 Stat Plus analyzer, YSI Incorporated, Yellow Springs, OH). The finger-stick blood samples were taken simultaneously with blood sample collection from the antecubital vein. For plasma epinephrine, blood samples were drawn into chilled tubes containing EGTA and glutathione (Amersham, RPN532 Vacutainer tubes), centrifuged, and the plasma stored at −80°C until analysis. Plasma concentrations of epinephrine were determined by high-pressure liquid chromatography (HPLC) with electrochemical detection (13). Plasma volume changes were estimated from blood hematocrit and hemoglobin values by using the method of Dill and Costill (7).

Statistical analysis.

Statistical significance was set at the P < 0.05 level, and values expressed as mean ± SE. Carbohydrate and placebo groups were compared for subject characteristics and race performance measures using Student’s t-tests (Tables 1 and 2). Heart rate and RPE measured during the marathon were analyzed using 2 (carbohydrate and placebo conditions) × 12 (times of measurement) repeated measures ANOVA. Hormone values were analyzed using 2 (carbohydrate and placebo groups) × 3 (times of measurement) repeated measures ANOVA. If the group × time interaction P-value was ≤ 0.05, the change from baseline to the immediate postrace value was compared between groups using Student’s t-tests. An independent t-test was used to determine differences in heart rate and RPE during the three measurements taken during the final 10 km.

Subject characteristics; mean ± SE.
Race performance measures; mean ± SE.


Ninety-eight of 102 runners including 12 females complied with all aspects of the study and finished the marathon race. Data from the 1999 Charlotte Marathon and 2000 Grandfather Mountain Marathon did not differ significantly and were therefore combined. Table 1 lists the subject characteristics for the carbohydrate (N = 48) and placebo (N = 50) groups. Age ranged from 21 to 72 yr. As a group, the marathon runners were highly experienced and committed to regular training and racing but were well below elite status as evidenced by their average personal record time of the previous year (just below 4 h for each group). The treadmill test data indicate a high degree of age-adjusted cardiorespiratory fitness. Ambient temperature and relative humidity were measured three times during each race event and averaged 19.1°C (range for both marathons was similar, 17.2–23.4°C) and 0.55 (0.45–0.65).

Race performance measures are contrasted between the carbohydrate and placebo groups in Table 2. Race times were slower than the marathoner’s personal record time of the previous year (Table 1) due to the hilly terrain of the Charlotte and Grandfather Mountain marathon race courses. Race time did not differ significantly between the carbohydrate and placebo groups (Table 2), but when contrasted with the runner’s personal record time of the previous year, the race time of the placebo group was 0.68 ± 0.05 h slower compared with 0.42 ± 0.06 h of the carbohydrate group (a 15.6-min differential) (P = 0.002). The beverage ingestion goal of 1 L·h−1 of either experimental or placebo during running was nearly met for each group, and as a result, body mass and plasma volume changes were slight (Table 2).

The heart rate and RPE data represent the mean of 12 measurements taken throughout the 42.2-km race events, and three measurements taken during the final 10 km. There was a trend for heart rate to be significantly different between carbohydrate and placebo groups [F (11,84)=1.85, P = 0.057] throughout the marathon (Fig. 1). Heart rate, when expressed as a percentage of the maximum heart rate, was lower with placebo (82.0% ± 0.6) than carbohydrate (84.2% ± 0.6) (P < 0.01) condition, especially during the final 10 km: placebo (78.7% ± 1.0) and carbohydrate (84.5% ± 0.7) (P < 0.001). In addition, heart rate (beat·min−1) was significantly different (P < 0.01) between groups during the final 10 km of the marathon: placebo (143 ± 2) and carbohydrate (152 ± 2) (P < 0.01).

The effect of carbohydrate or placebo ingestion on heart rate responses over the course of marathon running. * Significantly different means (P < 0.001) for heart rate (beat·min−1 or %HRmax) during the final 10 km between experimental conditions.

RPE increased over time for all subjects combined [F (11,84) = 41.5, P < 0.001]. RPE was not significantly different between carbohydrate and placebo ingestion [F (11,84) = 3.09, P = 0.08] throughout the marathon (Fig. 2). Average RPE throughout the race was not significantly different between the placebo (13.9 ± 0.2) and carbohydrate (13.4 ± 0.2) conditions (P = 0.08). There was a trend for average RPE during the final 10 km to be significantly different between conditions: placebo (16.8 ± 0.3) and carbohydrate (16.1 ± 0.3) (P = 0.06).

The effect of carbohydrate or placebo ingestion on ratings of perceived exertion over the course of marathon running. There was a trend for average RPE during the final 10 km to be significantly different between conditions: placebo (16.8 ± 0.3) and carbohydrate (16.1 ± 0.3) (P = 0.06).

The pattern of change in plasma glucose, insulin, lactate, and cortisol were significantly different between groups (Table 3). Postrace plasma glucose (P < 0.001), insulin (P < 0.001), and lactate (P < 0.05) levels were significantly lower in the placebo group, and postrace cortisol (P < 0.05) significantly higher in the placebo compared with carbohydrate group. The pattern of change in plasma catecholamines did not differ between the placebo and carbohydrate conditions.

Plasma glucose and hormonal changes in response to running a competitive marathon race in carbohydrate and placebo groups.


The purpose of this investigation was to examine the relation between ratings of perceived exertion (RPE) and levels of bodily carbohydrate substrate during a competitive marathon. This was accomplished by using a randomized and double-blind experimental design in which runners ingested either carbohydrate or placebo beverages while running at a self-selected velocity. RPE during running an actual marathon did not differ significantly between the carbohydrate and placebo conditions, although there was a nonsignificant trend toward a higher RPE during the later portion of the race with placebo. Despite the similarity in RPE between conditions, there was a significant decrease in plasma glucose and insulin concomitant with an increase in plasma cortisol and growth hormone with the placebo compared with the carbohydrate condition. In addition, the carbohydrate condition was associated with a significantly higher %HRmax during the later portion of the race.

We hypothesized that RPE would be lower during running with carbohydrate supplementation. However, such a change in RPE was not demonstrated as in our previous laboratory-based studies (12,20,21). A possible explanation for this finding is that the present study was conducted during an actual marathon race in which experimental outcomes could be easily affected by many extraneous variables, including weather, terrain, climate, and subject motivation. In addition, due to the fact that the intensity at which runners were performing at varied from point-to-point during the race, the magnitude of RPE could become responsive to the level of not only carbohydrate energy substrates but also relative cardiorespiratory and metabolic demand. Although not included in the present investigation, future research using RPE in field settings such as a competitive marathon might consider the employment of differentiated RPEs which may represent signals that are both peripheral (i.e., limbs) and respiratory-metabolic (i.e., chest).

Even though the present study did not find a significant difference in RPE between conditions during actual marathon running, the possibility of a physiological link between perception of exertion and carbohydrate energy substrate should not be ruled out. This assertion can be made because during the placebo condition there was a consistent but nonsignificant trend toward a higher RPE during the later portion of the race concomitant with lower plasma glucose values, in spite of the potential variability associated with a study of this nature. The physiological link may be plausible because during the carbohydrate condition where plasma glucose was well maintained, subjects were able to maintain a higher %HRmax at the same or even somewhat lower RPE during the later portion of running. It is likely that the intensity of perceived exertion would have been attenuated after carbohydrate supplementation if the subjects had run at the same or constant pace throughout the race in both the carbohydrate and placebo conditions.

In the previous laboratory studies (20,21), we have been able to measure the level of carbohydrate oxidation utilizing indirect calorimetry. Measurement of these data is important because it allows measurement of the utilization of bodily carbohydrate and its overall relation to perceived exertion. As the present study was conducted in a race/field setting, it was not possible to measure this metabolic parameter. However, the measurement of plasma cortisol and growth hormone may help provide an overall picture of the utilization of endogenous carbohydrate energy substrates during the marathon run. Cortisol and growth hormones are the glucoregulatory hormones that normally increase during the later stages of prolonged exercise when the level of endogenous carbohydrate decreases significantly (16). In the present study, plasma cortisol remained relatively unchanged with carbohydrate supplementation in contrast to an increase in the placebo condition. Although plasma growth hormone increased significantly, the magnitude of increase was much less during the carbohydrate as compared with placebo condition. These findings suggest that the attainment of a greater %HRmax at a given RPE can be attributable to a sustained supply of carbohydrate energy substrates to the exercising muscle.

Under the condition where the intensity of exercise varies as in competitive marathon running, it appears that factors other than carbohydrate energy substrate may be involved in mediating perceived exertion. This conclusion is derived from the fact that RPE did not differ significantly despite the differences in the level of blood glucose and glucoregulatory hormones between the carbohydrate and placebo condition. Previous studies have indicated that lactate threshold is a functional physiological anchor for the perception of exertion. These studies have observed that RPE at lactate threshold remains unaffected even though lactate threshold occurs at a higher cardiorespiratory and metabolic demand due to training (2,6,9,19). In this context, it is possible that the lack of differences in RPE found presently may have been due, at least in part, to the fact that our runners performed at a self-selected pace proximal to their lactate threshold for most of the race. This speculation is supported by the fact that the average RPE as shown in Table 2 was between 13 and 14 in both conditions and these figures are very similar to what was reported at lactate threshold in previous studies (6,9).

In conclusion, the present study is the first of its kind to examine the potential role that the availability of carbohydrate energy substrate plays in mediating the strength of perceived exertion during a marathon race. It was found that subjects undergoing carbohydrate supplementation were able to maintain their running at a higher %HRmax despite the fact that their perception of exertion was not significantly different from than placebo group. During prolonged strenuous exercise where intensity varies from point-to-point as in marathon running, it appears that factors other than carbohydrate energy substrate availability play an important role in mediating the strength of perceived exertion.


1. Borg, G. A. Psychophysical bases of perceived exertion. Med. Sci. Sports Exerc. 14: 377–381, 1982.
2. Boucher, S. H., R. L. Seip, R. K. Hetzler, E. F. Pierce, D. Snead, and A. Weltman. The effect of specificity of training on ratings of perceived exertion at the lactate threshold. Eur. J. Appl. Physiol. 59: 365–369, 1989.
3. Burgess, M. L., R. J. Robertson, J. M. Davis, and J. M. Norris. RPE, blood glucose, and carbohydrate oxidation during exercise: effects of glucose feedings. Med. Sci. Sports Exerc. 23: 353–359, 1991.
4. 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.
5. Coyle, E. F., A. R. Coggan, M. K. Hermert, and J. L. Ivy. Muscle glycogen utilization during prolonged strenuous exercise when fed carbohydrates. J. Appl. Physiol. 61: 165–172, 1986.
6. Demello, J. J., K. J. Cureton, R. E. Boineau, and M. M. Singh. Ratings of perceived exertion at the lactate threshold in trained and untrained men and women. Med. Sci. Sports Exerc. 19: 354–362, 1987.
7. 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.
8. Felig, P., A. Cherif, A. Minagawa, and J. Wahren. Hypoglycemia during prolonged exercise in normal men. N. Engl. J. Med. 306: 895–900, 1982.
9. Hill, D. W., K. J. Cureton, S. C. Grisham, and M. A. Collins. Effect of training on the rating of perceived exertion at the ventilatory threshold. Eur. J. Appl. Physiol. 56: 206–211, 1987.
10. Hyvarinen, A., and E. A. Nikkila. Specific determination of blood glucose with o-toluidine. Clin. Chim. Acta. 7: 140–143, 1962.
11. Ivy, J. L., D. L. Costill, W. J. Fink, and R. W. Lower. Influence of caffeine and carbohydrate feeding on endurance performance. Med. Sci. Sports Exerc. 11: 6–11, 1979.
12. Kang, J., R. J. Robertson, F. L. Goss, et al. Effect of carbohydrate substrate availability on ratings of perceived exertion during prolonged exercise of moderate intensity. Percept. Mot. Skills 82: 495–506, 1996.
13. Maruta, K., K. Fujita, S. Ito, and T. Nagatsu. Liquid chromatography of plasma catecholamines, with electrochemical detection, after treatment with boric acid gel. Clin. Chem. 30: 1271–1273, 1984.
14. Noakes, T. D., E. V. Lambert, M. I. Lambert, et al. Carbohydrate ingestion and muscle glycogen depletion during marathon and ultramarathon racing. Eur. J. Appl. Physiol. 57: 482–489, 1988.
15. Noble, B. J., R. J. Robertson. Peripheral and nonspecific physiological mediators. In:Perceived Exertion. Champaign, IL: Human Kinetics, 1996, pp. 125–155.
16. Powers, S. K., E. T. Howley. Hormonal responses to exercise. In:Exercise Physiology Textbook, 2nd Ed. Dubuque, IA: Brown & Benchmark, 1994, pp. 69–108.
17. Robertson, R. J., F. L. Goss, T. E. Auble, et al. Cross-modal exercise prescription at absolute and relative oxygen uptake using perceived exertion. Med. Sci. Sports Exerc. 22: 653–659, 1990.
18. Sherman, W. M., D. L. Costill, W. J. Fink, and J. M. Miller. Effect of exercise-diet manipulation on muscle muscle glycogen and its subsequent utilization during performance. Int. J. Sports Med. 2: 114–118, 1981.
19. Steed, J., G. A. Gaesser, and A. Weltman. Ratings of perceived exertion and blood lactate concentration during submaximal running. Med. Sci. Sports Exerc. 26: 797–803, 1994.
20. Utter, A. C., J. Kang, D. C. Nieman, et al. Effect of carbohydrate ingestion and hormonal responses on ratings of perceived exertion during prolonged cycling and running. Eur. J. Appl. Physiol. 80: 92–99, 1999.
21. Utter, A. C., J. Kang, D. C. Nieman, and B. Warren. Effect of carbohydrate substrate availability on ratings of perceived exertion during prolonged running. Int. J. Sports Nutr 7: 274–285, 1997.


© 2002 Lippincott Williams & Wilkins, Inc.