Secondary Logo

Journal Logo

Pre-exercise carbohydrate meals: application of glycemic index


Medicine & Science in Sports & Exercise: January 1999 - Volume 31 - Issue 1 - p 164-170
Applied Sciences: Physical Fitness And Performance

Pre-exercise carbohydrate meals: application of glycemic index. Med. Sci. Sports Exerc., Vol. 31, No. 1, pp. 164-170, 1999.

Purpose: The purpose of this study was to compare postprandial glycemic, insulinemic, and physiologic responses to a pre-exercise meal calculated to have a low glycemic index (LGI) with one calculated to have a moderately high glycemic index (HGI); each meal provided three foods totaling 1.5 g carbohydrate/kg body weight.

Methods: After an overnight fast, 10 trained cyclists consumed one of the test meals or water 30 min before cycling 2 h at 70% of maximum oxygen uptake (V˙O2max), followed by cycling to exhaustion at 100% of V˙O2max.

Results: Plasma insulin levels were significantly lower (P < 0.05) after LGI than after HGI through 20 min of exercise. Significantly higher (P < 0.05) respiratory exchange ratios were observed after HGI than after LGI until 2 h of exercise. At that time plasma glucose levels were significantly higher and ratings of perceived exertion lower (P < 0.05) after LGI compared with after HGI. Time to exhaustion was 59% longer after LGI (206.5 ± 43.5 s) than after HGI (129.5 ± 22.8 s).

Conclusions: These results suggest a pre-exercise LGI may positively affect maximal performance following sustained exercise. The LGI maintained higher plasma glucose levels at the end of 2 h of strenuous exercise than the HGI, which may have better supported subsequent maximal effort.

Department of Nutrition and Food Science and Department of Human Performance, San Jose State University, San Jose, CA 95192; and Veterans Affairs Health Care System, Palo Alto, CA 94304

Submitted for publication September 1996.

Accepted for publication January 1998.

The authors gratefully acknowledge the assistance of Sharon Moynihan, David Guido, Diane Cave, and the nursing staff of the Aging study unit at the Palo Alto Veterans Affairs Health Care System. In addition, we thank Ray Hintz for the insulin assays. We are also grateful to the subjects for their dedication and determination, without which this work could not have been accomplished.

Address for correspondence: Helen M. DeMarco, Department of Nutrition and Food Science, San Jose State University, One Washington Square, San Jose, CA 95192-0058.

Oxidation of plasma glucose increases during the latter stages of prolonged strenuous exercise when muscle glycogen nears depletion (10). The eventual lowering of plasma glucose levels may contribute to fatigue in exercising subjects (5). This response may be prevented by supplying an exogenous source of glucose to subjects during exercise (8,10). However, consuming carbohydrates during an athletic event is not always practical or possible. Alternatively, the maintenance of blood glucose during prolonged exercise may be accomplished through pre-exercise carbohydrate feedings, particularly low glycemic index foods (23), which are thought to enter the blood slowly.

The glycemic index (GI) is a ranking of foods based on their postprandial glycemic response (15). GI is equal to the postprandial change in blood glucose from a standard amount of a food divided by the postprandial change in blood glucose from a standard glucose load multiplied by 100. It has been suggested that the glycemic indices of mixed meals can be predicted from the glycemic indices of the component carbohydrate foods (4), with the weighted mean of the individual GI values based on the percentage of the total meal carbohydrate provided by each food (28). A positive correlation between the observed glycemic index and the predicted glycemic index of meals has been observed (4,6,28). In addition, postprandial insulin responses are important when studying physiological responses to food, and lower GI diets have been associated with reduced postprandial insulin secretions (4,16).

Researchers have recently studied the effects of pre-exercise carbohydrate feedings using the GI with isolated foods (22,23). In the latter part of endurance exercise, higher levels of glucose and free fatty acids (FFA) were observed after consumption of low GI foods than after consumption of high GI foods (22,23) and endurance time was longer following the low GI food than following the high GI food (23). The improved time is suggested to result because the low GI food induces less postprandial hyperglycemia and hyperinsulinemia, lower levels of plasma lactate before and during exercise, and the maintenance of plasma glucose and FFA at higher levels during critical periods of exercise.

The purpose of this study was to compare postprandial glycemic, insulinemic, and physiologic responses, as well as performance parameters, in trained cyclists to a meal with a calculated low GI index and one with a GI calculated to be moderately high. Assessment of the postprandial responses to meals designed based on the glycemic index may guide athletes in their choice of pre-exercise meals. It was hypothesized that given two meals with differing calculated GI, the meal with the lower GI would produce a lower insulin surge after consumption and would maintain higher blood glucose levels toward the end of 2 h of strenuous exercise; therefore, subsequent maximal exercise performance would be enhanced.

Back to Top | Article Outline


Subjects. Ten men, accustomed to cycling to fatigue over prolonged periods (>2 h), were recruited from local cycling teams. The training schedule was to remain constant throughout the period of study. Mean (± SD) age, weight, body fat, and maximum oxygen uptake (V˙O2max) were 30.7 ± 4.3 yr, 75.2 ± 7.0 kg, 7.7 ± 2.3%, and 61.2 ± 5.2 mL kg−1·min−1, respectively.

Before they were accepted into the study, subjects underwent an initial screening, which included health history, fasting blood analysis, urine analysis, a complete physical, and a resting electrocardiogram. The subjects were given 75 g of glucose (Trutol 100; Custom Laboratories, Inc., Baltimore, MD) and 2 h later an additional blood sample was drawn for determination of glucose tolerance. On a separate occasion, subjects performed a maximal exercise test using a cycle ergometer (model 818E, Monark, Varberg, Sweden). The load on the ergometer was increased at 3-min intervals while the subject continuously pedaled at 85 rpm. Cycling continued until the subject was unable to continue, even with encouragement. V˙O2max was established when at least two of the three following criteria were met: 1) a leveling off of V˙O2 with increasing workload, 2) a heart rate (HR) within 10 beats·min−1 of predicted maximum, and 3) a respiratory exchange ratio (RER) ≥ 1.1. Expired air was continuously analyzed for oxygen and carbon dioxide using a metabolic cart (Ametek, Pittsburgh, PA). The information gained from this test was used to determine the intensity of exercise for the experimental trials of the study.

The protocol for the study was approved by the Institutional Committee for the Protection of Human Subjects at Stanford University and the Human Subjects-Institutional Review Board at San Jose State University. Written informed consent was obtained from all subjects before participation in the study.

Study design. Subjects participated in three exercise trials. One of three meals, randomly assigned, was consumed 30 min before cycling for 2 h at 70% of V˙O2max. This exercise intensity would be applicable to competitive cycling (14). This endurance period was followed by a performance trial, where subjects were required to cycle to exhaustion at 100% of V˙O2max. The tests were conducted approximately 1 wk apart. Subjects arrived at the metabolic unit in the morning after a 12-h overnight fast, and a catheter was inserted into an antecubital vein. After an initial blood sample (0 time sample), the subjects consumed the pre-exercise meal. Fifteen and 30 min after completing the meal, blood samples were drawn; during this time the subject remained seated and rested. Cycling commenced immediately after the 30-min sampling. Throughout all trials, the load on the cycle ergometer and the rate of pedaling were carefully monitored. Oxygen consumption, RER, ratings of perceived exertion (RPE) based on Borg's 20 point scale (3), and blood samples were obtained every 20 min during the first 2 h of exercise. At the end of 2 h, subjects completed the performance trial. Subjects were strongly encouraged to continue and were blinded to time. The end point was defined as the point of volitional exhaustion or when pedaling dropped below 80 rpm because of fatigue. Oxygen consumption and RER were measured continuously throughout the performance trial, and a final blood sample was taken immediately at exhaustion.

For 2 d before each trial, subjects did not engage in strenuous exercise and maintained a standardized carbohydrate intake (6-8 g·kg−1 body wt). By maintaining a constant diet and training schedule, within-subject variation in glycogen levels before exercise should have been minimized. This protocol has been shown to give reproducible amounts of glycogen stores (10).

Exercise conditions. Dehydration was minimized by having the subjects drink at least 150 mL of water every 20 min (after each blood sampling). In addition, subjects were fan cooled throughout exercise. All rides were performed in a mean (± SD) room temperature of 21 ± 1°C.

Test meals. Two of the three meals tested included a combination of three carbohydrate foods designed to have different glycemic indices. Cereal, fruit, and dairy contributions to the total carbohydrate in the meal were 55%, 30%, and 15%, respectively. The moderately high glycemic index meal (HGI) included Cornflakes (Kellogg Co., Battle Creek, MI), banana, and milk (meal GI = 69.3). The low glycemic index meal (LGI) included All-Bran (Kellogg Co.), apple, and unsweetened yogurt (meal GI = 36). The third meal served as the control (CON) and included only water. The glycemic index of each meal (Table 1) was calculated according to the method of Wolever et al. (28). The amount of carbohydrate in the two meals was adjusted for each subject to provide 1.5 g of available carbohydrate per kg body weight. The types of carbohydrate in the two meals were the same, with starch, fructose, and lactose being the primary carbohydrate type for the cereal, fruit, and dairy, respectively. The fiber, fat, and protein content of the two test meals varied, with higher amounts of these components being in the LGI (Table 2). Water was consumed with meals as necessary to standardize the total volume of meals to minimize differences in gastric emptying.





Blood sampling and analysis. Five milliliters of blood was obtained at each sampling; lithium heparin was used as the anticoagulant. Blood samples were centrifuged at 3000 rpm for 10 min at 5°C, and separated plasma glucose was promptly analyzed by the glucose oxidase method (Beckman Instruments, Brea, CA). The remaining plasma was frozen at −70°C and analyzed later for insulin using radioimmunoassay (Diagnostic Systems Laboratories Inc., Webster, TX).

Statistical analysis. Results are expressed as mean ± SE. The data from the three trials were analyzed using a two-way ANOVA for repeated measures (meal × time). The computer program SPSSX (SPSS Inc., Chicago, IL) was used. When ANOVA indicated significant main effects, Tukey post-hoc tests were used to isolate significantly different means (P < 0.05).

Back to Top | Article Outline


Oxygen consumption and percent V˙O2max remained stable throughout the 2-h endurance period and were similar among trials. Plasma glucose levels (Fig. 1) were significantly higher in both the HGI and LGI trials at 15 min after the ingestion of the meals compared with those in the CON trial. By 30 min postmeal, the differences had been eliminated. At 15 and 30 min postmeal, plasma glucose levels were higher in the HGI trial than the LGI trial; however, the differences were not significant. During the CON trial, plasma glucose levels were maintained at the onset of exercise. In contrast, there was a sharp decline in plasma glucose concentrations in HGI and LGI 20 min into the exercise. These levels were significantly lower than the CON trial. The HGI trial, which demonstrated the greatest glycemic response before exercise (6.6 ± 0.4 mmol·L−1), had the greatest absolute decline in blood glucose concentration during the first 20 min of exercise (3.8 ± 0.2 mmol·L−1). The LGI meal resulted in a more gradual rise and fall in plasma glucose than that seen after the HGI meal, although the nadir reached by each meal was essentially the same. From 40 min of exercise until 100 min of exercise, there were no significant differences in plasma glucose concentrations among the three trials. However, by 120 min of exercise, plasma glucose concentrations for HGI and CON trials had decreased significantly from 40 min of exercise, while the concentration for LGI trial was still being maintained. At 120 min of exercise, plasma glucose levels were significantly higher in the LGI trial compared with those in the HGI and CON trials. After the performance trial (exhaustion), plasma glucose levels were significantly higher in the LGI trial compared with those in the HGI and CON trials. Moreover, at exhaustion, the plasma glucose levels for the HGI trial (4.1 ± 0.2 mmol·L−1) and the CON trial (4.2 ± 0.2 mmol·L−1) were significantly lower than prefeeding baseline levels, whereas the LGI trial plasma glucose level (4.6 ± 0.2 mmol·L−1) was still being maintained near baseline.

Figure 1

Figure 1

Plasma insulin levels increased significantly above baseline values 15 and 30 min after the HGI and LGI meals (Fig. 2). However, the rise was smaller after the LGI meal compared with the rise after the HGI meal. At 15 min and 30 min postmeal, plasma insulin levels in the LGI trial were significantly lower (35% and 45% less, respectively) than in the HGI trial. By 20 min of exercise, insulin levels had significantly decreased in both the HGI and LGI trials; yet plasma insulin levels in the HGI trial were still significantly higher compared with those in the LGI and CON trials. By 60 min of exercise, the difference in plasma insulin levels between the HGI and LGI trials was eliminated; yet both insulin levels were still significantly higher compared with those in the CON trial. By 120 min of exercise, differences among trials were eliminated.

Figure 2

Figure 2

Respiratory exchange ratios (RER) are shown in Figure 3. For all trials the RER was highest at the onset of exercise and gradually decreased throughout exercise. The RER during the first 100 min of exercise were significantly higher in the HGI trial compared with those in the LGI and CON trials. However, between 100 and 120 min of exercise, there was a significant decline in RER for the HGI trial, and at 120 min of exercise, there were no significant differences among trials. At no time during exercise were the differences between the LGI and CON trials significant.

Figure 3

Figure 3

Ratings of perceived exertion (RPE) increased gradually over time (Fig. 4). However, in all trials the increase in RPE did not become significant until 60 min of exercise. Only in the HGI and CON trials did RPE increase significantly beyond that point. At all times throughout exercise, RPE for the LGI trial were lower than for the HGI trials, although the differences were only significant at 20, 60, 80, and 120 min of exercise. At no time throughout exercise were there significant differences between the RPE for the HGI and CON trials.

Figure 4

Figure 4

Times to exhaustion for the performance trial are shown in Figure 5. Time to exhaustion in the LGI trial (206.5 ± 43.5 s) was significantly longer than in the HGI (129.5 ± 22.8 s) and the CON (120.0 ± 31.4 s) trials. V˙O2max, during the performance ride, was also significantly higher in the LGI trial (59.4 ± 2.08 mL·kg−1 ·min−1) than in the HGI (50.4 ± 2.07 mL·kg−1·min−1) and the CON (48.5 ± 3.11 mL·kg−1·min−1). There were no significant differences in both time to exhaustion and V˙O2max between the HGI and CON trials.

Figure 5

Figure 5

Back to Top | Article Outline


The purpose of this study was to compare metabolic and physiologic responses in trained cyclists to two mixed carbohydrate meals of differing calculated GIs. In addition, performance parameters were compared. The LGI produced a lower insulin surge after consumption than did the HGI. In addition, blood glucose levels were higher and ratings of perceived exertion were lower toward the end of 2 h of exercise in the LGI trial. Subsequently, in the performance trial, the LGI trial resulted in a longer time to exhaustion. Many competitive sports begin in the early morning to minimize the possible decrements in performance associated with elevated environmental temperature. Choosing appropriate morning meals to optimize energy availability during the event can be an important factor in enhancing performance. These results provide support for the hypothesis that a meal with a calculated LGI may confer an advantage over one with a calculated HGI consumed before prolonged strenuous exercise.

The GI was developed by Jenkins et al. (15) to classify the effect of different carbohydrate foods on blood glucose. Based on previous data (4,6,28), we thought using calculated GIs of mixed meals would allow us to predict the postprandial glycemic effects of meals. However, in the present study we did not find significant differences between the glycemic responses to the two meals. This lack of difference may result from rapid postprandial glucose disposal as a result of enhanced insulin sensitivity observed in trained subjects (19). This enhanced insulin sensitivity may result from several factors, among them an increase in muscular mass (29) and increases in muscular blood flow and glucose uptake (11), as well as an increased cellular metabolism of glucose (11). As the GI was constructed primarily using postprandial responses observed in untrained subjects, the responses in trained subjects would be expected to be less dramatic. The insulinemic responses paralleled the pattern of the glycemic responses but resulted in values that were significantly different between meals. It could be that the insulin responses were a reflection of a greater variation in the glycemic responses between meals before the 30-min postmeal blood sampling.

Since our goal was to test the effect of real foods, the noncarbohydrate components of the meals varied. The fat, protein, and total energy content were slightly higher in the LGI as compared with that in the HGI. Fat and protein may affect the glycemic response by delaying gastric emptying rate (27) and/or acting as insulin secretagogues (4). One might predict there would have been an increase in insulin secretion following the LGI, as protein consumed in conjunction with a carbohydrate load increases insulin secretion (18). That the amount of protein required to have significant effects on postprandial glycemic and insulinemic levels would have to be large (18) may explain the lack of effect of protein seen here. The energy difference between meals (15%) was small and since the energy density of meals may not be related to the rate of gastric emptying (17), the possibility that energy differences may have influenced postprandial effects of the meals seems slight.

The factor most likely to influence the effects of the meals in this study was the fiber content, as it varied dramatically between meals (57 g in LGI vs 5 g in HGI). The blunted glycemic response that is typically seen following consumption of high fiber foods is thought to be caused by a reduced rate of gastric emptying and a reduced rate of intestinal absorption (24). In addition, it has been observed that phytic acid, a component associated with wheat bran found in the All-Bran cereal, can reduce the rate of digestion in vitro and the glycemic response in vivo(30). Therefore, the reduced glycemic response in the LGI, although not statistically different from the HGI, was most likely caused by the differences in the fiber content between meals.

Elevated postprandial plasma insulin concentrations in the HGI and LGI trials were primarily responsible for the rapid decline in plasma glucose levels during the first 20 min of exercise (7). Insulin is also responsible for the increased rate of carbohydrate oxidation typically seen after consumption of carbohydrates compared with that in the fasted state (9,12,14,22). Based on the RER values, the rate of carbohydrate oxidation in the HGI trial remained significantly higher than in the other trials until 120 min of exercise, despite plasma insulin levels returning to baseline values. It has been shown that the effects of high insulin levels can persist for extended periods after plasma insulin levels have dropped, with an inhibition of adipocyte lipolysis even with elevated levels of epinephrine (21). The lower RER value in the LGI trial suggests a greater percentage of fat oxidation during exercise than in the HGI trial and possibly increased availability of free fatty acids (FFA) as an energy source to the exercising muscle. When meals with significant differences in insulinemic responses are compared, significant differences in plasma levels of FFA and glycerol are found (22), and enhanced FFA availability and utilization may bring about glycogen sparing (7). In the present study, the LGI meal induced lower plasma insulin levels and a lower RER throughout the most of the exercise, and although neither FAA nor muscle glycogen were measured, the possibility that muscle glycogen stores may have been spared cannot be completely dismissed. Costill et al. (7) observed a significant increase in muscle glycogen utilization following the ingestion of 75 g of glucose 45 min before exercise, and intense exercise performance can be reduced in exercising subjects with lowered muscle glycogen content (1).

Additionally, recent studies have shown that the observed improvements in performance following ingestion of various carbohydrates may result from the maintenance of blood glucose late in exercise (10,23). Few studies have directly examined the effects of varying the GI of pre-exercise foods. Thomas et al. (22) found that lentils (GI = 29) maintained plasma glucose levels at physiologically favorable rates for 90 min of exercise (24) and resulted in a longer time to exhaustion than glucose and potatoes (GI = 100 and GI = 98, respectively). In a subsequent study, Thomas et al. (23) found that plasma glucose levels toward the end of prolonged strenuous exercise correlated inversely with the GI of the food consumed before exercise; a 10-unit difference in GI was associated with a 0.2 mmol·L−1 difference in plasma glucose concentration at the end of exercise. The test meals consisted of two high GI foods (potato flake meal and rice cereal meal) and two low GI foods (lentil flakes meal and bran cereal meal), each providing 1 g carbohydrate per kilogram body weight. These test meals were similar to those used in the current study in regards to protein and fat distribution, although the energy content in our meals was higher. Our results are similar to the results of Thomas et al. in that we found a 10-unit difference in GI was associated with about a 0.15 mmol·L−1 difference in plasma glucose concentration at the end of 2 h of exercise. Low GI carbohydrates theoretically release glucose from the gut for extended periods, and by maintaining blood glucose levels over a longer period of time, hepatic glycogen stores may also be spared. On the other hand, the pre-exercise elevation in insulin in the HGI trial may have increased muscle glucose uptake during exercise, with the eventual lowering of plasma glucose as intestinal release of glucose and/or hepatic glucose production declined.

Increases in blood glucose uptake by the exercising muscles may be accelerated when muscle glycogen concentrations are reduced (2). Therefore, lowering of blood glucose during the latter stages of exercise can contribute to the development of fatigue if leg glucose uptake cannot offset the reduced muscle glycogen availability (10). However, since glucose uptake by the legs may increase to almost 8 mmol·min−1 during maximal exercise (26) and since the decrease in plasma glucose levels was not large in the performance trial, the differences in plasma glucose levels among trials may only partially account for the performance differences observed.

In addition, the subjects never experienced hypoglycemia (plasma glucose < 3.5 mmol·L−1) throughout the trials. Yet, immediately before the performance ride and at exhaustion, plasma glucose levels were significantly lower in the HGI trial than the LGI trial. It has been observed that some athletes experience fatigue when plasma glucose levels are moderately reduced, in the absence of hypoglycemia (8). The significant decline in the RER in the HGI trial at the end of 2 h of exercise indicated that carbohydrate utilization, and possibly carbohydrate availability, was abating. At the same time, ratings of perceived effort rose. As sources of carbohydrate become depleted, ratings of perceived exertion rise and inevitably the intensity of exercise most be reduced or exercise discontinued (20).

In conclusion, a significant improvement in maximal exercise performance time, following endurance exercise, was observed when a calculated LGI meal was consumed 30 min before exercise as compared with the consumption of a meal calculated to have a HGI. These results may support the use of low GI meals for individuals participating in early morning competitive events. Increasing the availability of blood glucose exogenously via a slow-releasing source of glucose may supplement endogenous stores adequately enough to enhance performance during maximal exercise. Nevertheless, further research is needed, testing a number of mixed carbohydrate meals of differing computed GIs to assure that the results observed in the present study result from the GI itself rather than from any potential uniqueness of the two test meals.

Back to Top | Article Outline


1. Akermark, C., I. Jacobs, M. Rasmusson, and J. Karlsson. Diet and muscle glycogen concentration in relation to physical performance in Swedish elite ice hockey players. Int. J. Sport Nutr. 6:272-284, 1996.
2. Bonen, A., S. A. Malcolm, R. D. Kilgour, K. P. MacIntyre, and A. N. Belcastro. Glucose ingestion before and during intense exercise. J. Appl. Physiol. 50:766-771, 1981.
3. Borg, G. A. V. Psychophysical basis of physical exertion. Med. Sci. Sports Exerc. 14:377-387, 1982.
4. Chew, I., J. C. Brand, A. W. Thorburn, and A. S. Truswell. Application of glycemic index to mixed meals. Am. J. Clin. Nutr. 47:53-56, 1988.
5. 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.
6. Collier, G. R., T. M. S. Wolever, G. S. Wong, and R G. Josse. Prediction of glycemic response to mixed meals in non-insulin-dependent diabetic subjects. Am. J. Clin. Nutr. 44:349-353, 1986.
7. Costill, D. L., E. Coyle, W. Dalsky, W. Fink, and D. Hoopes. Effects of elevated plasma FFA and insulin on muscle glycogen usage during exercise. J. Appl. Physiol. 43:695-699, 1977.
8. 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.
9. Coyle, E. F., A. R. Coggan, M. K. Hemmert, R. C. Lowe, and T. J. Walters. Substrate usage during prolonged exercise following a pre-exercise meal. J. Appl. Physiol. 59:429-433, 1985.
10. 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.
11. Ebeling, P., R. Bourey, L. Koranyi, et al. Mechanism of enhanced insulin sensitivity in athletes: increased blood flow, muscle glucose transport protein (GLUT-4) concentration, and glycogen synthase activity. J. Clin. Invest. 92:1623-1631, 1993.
12. Foster, C., D. L. Costill, and W. J. Fink. Effects of pre-exercise feedings on endurance performance. Med. Sci. Sports Exerc. 11:1-5, 1979.
13. Foster-Powell, K. and J. Brand Miller. International tables of glycemic index. Am. J. Clin. Nutr. 62:871S-893S, 1995.
14. Horowitz, J. F. and E. F. Coyle. Metabolic responses to pre-exercise meals containing various carbohydrates and fat. Am. J. Clin. Nutr. 58:235-241, 1993.
15. Jenkins, D. J. A., T. M. S. Wolever, R. H. Taylor, et al. Glycemic index of foods: a physiological basis for carbohydrate exchange. Am. J. Clin. Nutr. 34:362-366, 1981.
16. Jenkins, D. J. A., T. M. S. Wolever, G. R. Collier, et al. Metabolic effects of a low-glycemic-index diet. Am. J. Clin. Nutr. 46:968-975, 1987.
17. Mourot, J., P. Thouvenot, C. Couet, J. M. Antoine, A. Krobicka, and G. Debry. Relationship between the rate of gastric emptying and glucose and insulin responses to starchy foods in young healthy adults. Am. J. Clin. Nutr. 48:1035-1040, 1988.
18. Nuttall, F. Q., A. D. Mooradian, M. C. Gannon, C. Billington, and P. Krezowski. Effect of protein ingestion on the glucose and insulin response to a standardized oral glucose load. Diabetes Care 7:465-470, 1984.
19. Rodnick, K. J., W. L. Haskill, A. L. M. Swislocki, J. E. Fley, and G. M. Reaven. Improved insulin action in muscle, liver, and adipose tissue in physically trained human subjects. Am. J. Physiol. 253:E489-E495, 1987.
20. Sherman, W. M., M. C. Peden, and D. A. Wright. Carbohydrate feedings 1 h before exercise improves cycling performance. Am. J. Clin. Nutr. 54:866-870, 1991.
21. Solomon, S. S. and W. C. Duckworth. Effect of antecedent hormone administration on lipolysis in the perfused isolated fat cell. J. Lab. Clin. Med. 88:984-994, 1976.
22. Thomas, D. E., J. R. Brotherhood, and J. C. Brand. Carbohydrate feeding before exercise: effect of glycemic index. Int. J. Sports Med. 12:180-186, 1991.
23. Thomas, D. E., J. R. Brotherhood, and J. Brand Miller. Plasma glucose levels after prolonged strenuous exercise correlate inversely with glycemic response to food consumed before exercise. Int. J. Sport Nutr. 4:361-373, 1994.
24. Thorne, M. J., L. U. Thompson, and D. J. A. Jenkins. Factors affecting starch digestibility and the glycemic response with special reference to legumes. Am. J. Clin. Nutr. 38:481-488, 1983.
25. United States Department of Agriculture. Home and Garden Bulletin, No. 72. Nutritive Value of Foods. Washington, D.C.: U.S. Government Printing Office, 1988.
    26. Wahren, J., P. Felig, G. Ahlborg, and L. Jorfeldt. Glucose metabolism during leg exercise in man. J. Clin. Invest. 50:2715-2725, 1971.
    27. Welsh, I. M. L., C. Bruce, S. E. Hill, and N. W. Read. Duodenal and ileal lipid suppresses postprandial blood glucose and insulin responses in man: possible implications for the dietary management of diabetes mellitus. Clin. Sci. 72:209-216, 1987.
    28. Wolever, T. M. S., D. J. A. Jenkins, A. L. Jenkins, and R. G. Josse. The glycemic index: methodology and clinical implications. Am. J. Clin. Nutr. 54:846-854, 1991.
    29. Yki-Jarvinen, H. and V. A. Koivisto. Effect of body composition on insulin sensitivity. Diabetes 32:965-969, 1983.
    30. Yoon, J. H., L. U. Thompson, and D. J. A. Jenkins. The effect of phytic acid on in vitro rate of starch digestibility and blood glucose response. Am. J. Clin. Nutr. 38:835-842, 1983.


    © 1999 Lippincott Williams & Wilkins, Inc.