INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE INTAKE OF A LARGE carbohydrate (CHO)-rich meal (>200 g CHO) 1-4 h before prolonged (>90 min), submaximal exercise [~70% of maximal oxygen uptake (O_{2 max})] increases cycling time to exhaustion (35) and enhances work output (27) or time trial performance (31) undertaken at the end of a standardized exercise bout. Preexercise CHO intake may be important for maintaining blood glucose concentrations via hepatic glucose output during the latter stages of such events. However, the elevation of plasma insulin concentration following such CHO feedings presents a potential disadvantage, in that it may suppress fat metabolism (14, 24), increase CHO oxidation (14, 17), and cause a decline in plasma glucose concentration (5, 10, 12, 16, 17, 24) during subsequent exercise. These metabolic disturbances may be attenuated by choosing preevent CHO sources that produce a minimal glycemic and insulinemic response (10, 15-17).

The application of the glycemic index (GI) of CHO-rich foods to sports nutrition was first investigated by Thomas and co-workers (33). They found, in comparison to a pretrial meal of a high-GI (HGI) food (potatoes), that exercise time to exhaustion was longer when a preexercise meal of a low-GI (LGI) CHO-rich food (lentils) was consumed 1 h before cycling at 67% of O_{2 max}. These findings have been widely publicized and have resulted in popular advice that athletes should choose preexercise meals based on LGI CHO-rich foods and drinks (4).

However, in endurance exercise, the most effective and commonly used strategy by athletes to promote CHO availability is to ingest CHO-rich drinks or foods during the event. To date, the interaction of preexercise and during-exercise CHO intake has received only brief attention (35). Therefore, the purpose of this investigation was to examine whether the GI of preexercise CHO intake has any impact on exercise metabolism and subsequent performance when large amounts of CHO are consumed during the exercise session. The nutritional strategies employed before and during the exercise trial were chosen to balance current recommendations for endurance athletes with realistic practices. A cycling protocol that allowed a relatively reliable and sports-specific measurement of performance was employed.

	SUBJECTS AND METHODS

Top Abstract Introduction Methods Results Discussion References

Subjects and preliminary testing. Six endurance-trained male cyclists (age 22.8 ± 2.3 yr, weight 72.0 ± 4.1 kg, O_{2 max} 68.6 ± 3.8 ml · kg¹ · min¹) volunteered to participate in this investigation, which was approved by the Research and Ethics Committee of the Faculty of Medicine of the University of Cape Town, South Africa. Because tracer amounts of [U-¹⁴C]glucose were ingested and blood samples were taken, the risks were carefully explained to the subjects before their written consent was obtained. The total radiation dose received by each subject was ~20 mrem. The radiation dose accepted as safe in South Africa is 500 mrem/yr or 130 mrem/13 wk (3).

Subjects were tested for O_{2 max} on an electronically braked cycle ergometer (Lode, Groningen, The Netherlands) modified with clip-on pedals and racing handlebars. The incremental cycle test to exhaustion and the accompanying gas-collection procedures have been described in detail previously (20). Briefly, each subject started cycling at an exercise intensity of 3.33 W/kg body mass (BM) for 150 s, after which the work rate was increased by 50 W for a further 150 s. Thereafter, the exercise intensity was increased by 25 W every 150 s up to the point of exhaustion.

The results of the initial maximal test were used to determine the exercise intensity that corresponded to 70% of each subject's O_{2 max} for use in the three subsequently described experimental rides. Throughout the maximal test, subjects wore a face mask attached to an Oxycon Alpha automated gas analyzer (Jaeger, The Netherlands). Before each test, the gas analyzer was calibrated by using a Hans Rudolf 5530 3-liter syringe and a 5% CO₂-95% N₂ gas mixture. Analyzer outputs were processed by an IBM computer, which calculated minute ventilation, oxygen consumption (O₂), and rates of CO₂ production (CO₂) using conventional equations.

Study design. Each subject undertook three trials in a randomized counterbalanced order, with each trial separated by a period of 7 days. On each occasion, one of the following test meals was consumed 2 h before cycling at 70% O_{2 max}: an HGI CHO-rich meal (HGI trial) of instant mashed potato (GI = 87, where glucose = 100; Ref. 13), an LGI CHO-rich meal (LGI trial) of lightly cooked pasta (GI = 37, Ref. 13), or a control meal (Con trial) of low-energy jelly. The CHO-rich meals provided 2 g CHO/kg BM, and the water content of all meals was standardized so that each provided ~1,100 ml of fluid (Table 1).

Food intake and training were standardized for the 24-h period before each trial. Subjects were provided with guidelines for CHO-rich meals and were requested to record their dietary intake on the day before the first trial. Identical food was consumed before the subsequent trials, with dietary records being continued to check compliance. Additionally, subjects were asked to abstain from alcohol, caffeine, and strenuous exercise for at least 24 h before each trial. On the morning of an experiment, subjects reported to the laboratory in an overnight-fasted state, and their food and training diaries were examined to ensure that all instructions had been followed. Thereafter, a flexible 18-gauge catheter was inserted into a forearm vein and attached to a three-way stopcock for the sampling of blood. The catheter was kept patent throughout the experiment with periodic injection of heparinized saline.

After a fasting blood sample was taken, the subjects were given 15 min to consume their test meal and rested for 2 h. Fifteen minutes before exercise, the subjects consumed 4 ml/kg BM of a 10 g/100 ml [U-¹⁴C]glucose solution (Amersham International, Buckinghamshire, UK) with a specific radioactivity of 0.17 µCi/g (6.3 kBq/g). The [U-¹⁴C]glucose label was added to the drink so that the rates of ingested glucose oxidation could be calculated. A total of 3.3 ml/kg BM of the labeled drink was ingested every 20 min during the steady-state ride, for a total of 24 ml/kg each trial (2.4 g of CHO/kg BM).

Immediately before the start of exercise, subjects voided, were weighed, and the cycle ergometer was set up to conform with their preferred cycling position. Exactly 2 h postprandial, the subjects started their steady-state ride at ~70% of O_{2 max} (245 ± 18 W). As part of the 2-h cycle, a warm-up period was allowed; exercise started at 100 W for 5 min, after which the workload was increased by 50 W/min until the final workload was attained. During the trials, the subjects were cooled with an electric fan while the laboratory was maintained at a constant temperature (~20°C) and relative humidity (~55%). On completion of the 2-h steady-state ride, the subjects were given 1 min to rest before they started the performance ride, which consisted of the time to complete 300 kJ. During the timed ride, subjects were kept blind to time; the only feedback given was the completion of each 50 kJ of work. After subjects had completed 275 kJ, they received information about each successive 5 kJ until the end of the ride. No performance results were provided to any subject until the completion of the entire study

Blood sampling and analysis. A fasting blood sample was collected before ingestion of the experimental meal whereafter blood sampling was undertaken, 30, 60, 90, and 120 min postprandially. During the steady-state ride, blood was sampled at successive 20-min intervals, commencing after 20 min of the start of the ride. Approximately 6 ml of blood were drawn at each sampling, of which 2 ml were placed in a tube containing potassium oxalate and sodium fluoride for the later analysis of plasma glucose concentrations. The remaining 4 ml were placed in a tube containing gel and clot activator and allowed to clot for 15 min at room temperature for the later analyses of serum insulin and free fatty acid (FFA) concentrations. All samples were kept on ice during the duration of the trial before the plasma and serum were separated by centrifugation (2,000 revolutions/min) at 4°C and stored at 18°C until subsequent analyses.

Plasma glucose concentrations were determined by the glucose oxidase method (Glucose Analyzer 2, Beckman Instruments, Fullerton, CA). Serum insulin concentrations were determined by the use of a commercially available radioimmunoassay kit (Count-A-Coat Insulin, Diagnostic Products). Serum total FFA concentrations were determined by an enzymatic colorimetric assay (Half-micro test, Boehringer Mannheim, Germany).

O₂, CO₂, and ¹⁴CO₂ measurements. Immediately after each blood sampling during the steady-state ride, gas exchange (O₂, CO₂) was measured for 5 min. In addition, CO₂ was trapped by passing a sample of expired air, collected in an ambulatory bag, through a solution containing 1 ml of 1 N hyamine hydroxide in methanol (United Technologies, Packard, IL), 1 ml of 96% ethanol (SAARCHEM, Krugersdorp, South Africa), and 2-3 drops of phenolphthalein (SAARCHEM). The expired air was bubbled through the trapping mixture until the solution became clear, at which point exactly 1 mmol of CO₂ had been absorbed (30). Liquid-scintillation cocktail (Ready Gel, Beckman Instruments) was added, and ¹⁴CO₂ radioactivity (in dpm) was counted in an Insorb 460C automatic liquid-scintillation counter (United Technologies).

(1)

(2)

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Records kept by each subject during the 24 h before each trial indicated compliance with the standardized preparation protocol; reported CHO intakes on the day before the three trials were 479 ± 51, 465 ± 62, and 460 ± 52 g for the HGI, LGI, and Con trials, respectively (not significant), and all subjects refrained from exercise during that day.

Figure 1 shows plasma glucose, serum insulin, and serum FFA concentrations for the 2-h period following ingestion of each test meal and for the subsequent 120 min of steady-state exercise. There was a significant interaction of diet and time for all parameters (P < 0.05). Thirty minutes after ingestion of the HGI meal, plasma glucose concentrations were significantly increased above fasting values (Fig. 1A). At this time point, plasma glucose concentrations were greater in both HGI and LGI trials (7.9 ± 0.6 and 6.6 ± 0.5 mmol/l) than in the Con trial (4.4 ± 0.3 mmol/l). Plasma glucose concentrations returned to baseline 60 min after each of the CHO meals but did not fall below fasting values in any of the trials. Plasma glucose concentrations rose slightly with the ingestion of the CHO drink (starting 15 min before the start of exercise) and at the onset of exercise, and remained at euglycemic levels (4.5-6.0 mmol/l) throughout the steady-state ride in all trials. Plasma glucose concentrations during exercise did not differ among trials (Fig. 1A).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Plasma glucose (A), serum insulin (B), and serum free fatty acid (FFA; C) concentrations (marked by brackets) for the 2-h period postprandial and during 2 h of steady-state exercise. Values are means ± SE for 6 subjects. HGI, high-glycemic index trial; LGI, low-glycemic index trial; Con, control trial. * Significantly different from Con; # significantly different from HGI; a significantly different from fasting; b significantly different from start of exercise (P < 0.05 for all symbols).

The rise in serum insulin after the HGI meal was significantly greater and persisted for longer than for the LGI meal (Fig 1B). Serum insulin concentrations at 30 min were higher in the HGI trial than in the LGI trial (74.1 ± 10.1 vs. 43.9 ± 8.9 uU/ml) and remained elevated above those in the LGI trial at 60 and 90 min. Serum insulin concentrations remained at fasting values (<10 uU/ml) in the Con trial until 90 min; however, at 120 min, there was an increase in insulin concentration in response to the ingestion of the glucose drink 15 min before the start of exercise. In the LGI trial, serum insulin concentration peaked at 30 min after the meal and then declined thereafter, with the feeding of the bolus of the glucose drink causing a small rise. During the bout of steady-state exercise, despite the intake of significant amounts of CHO, serum insulin fell to fasting concentrations in all trials. Insulin concentrations in the HGI trial were still elevated above those in the Con trial at 20 min of exercise; thereafter, there were no differences in the serum insulin concentrations during exercise among any of the trials.

The intake of CHO from both the HGI and LGI meals caused a suppression of serum FFA concentrations to <0.1 mmol/l, which persisted until the start of the exercise bout (Fig. 1C). During the Con trial, FFA concentrations were maintained at fasting levels throughout the preexercise phase. However, intake of CHO at the onset of exercise caused a small drop in FFA concentration after 20 min of exercise. Nevertheless, at this time point, FFA concentrations in the Con trial were significantly greater than in the HGI trial. Thereafter during exercise, despite the continued intake of CHO, there was a gradual rise in serum FFA in all trials. After 120 min of exercise in the HGI and LGI trial, FFA concentrations were significantly greater than at the beginning of exercise, and in all trials FFA concentrations had returned to fasting values (~0.25 mmol/l) by this time. From 20 min of exercise onward, there were no differences in FFA concentrations among the three trials.

Metabolic data from 120 min of steady-state exercise are presented in Fig. 2. There was no significant interaction of diet and time for RER (Fig. 2A), total CHO oxidation (Fig. 2B), and oxidation of the ingested glucose drink (Fig. 2C). The decline in RER over the 120 min of steady-state exercise was not significant. Oxidation of CHO from the glucose drink increased throughout exercise from ~0.3 g/min after 20 min to ~0.8 g/min at 120 min across all trials. After 20 min, oxidation of the ingested CHO drink was significantly lower in the LGI trial than in Con; however, there were no differences at any other time points. Overall, the drink provided ~16% of total CHO oxidation for the 120 min of exercise, and this contribution was similar for all trials (Fig. 3). Total CHO oxidation was ~380 g during the 120 min of steady-state exercise and did not differ among trials.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Respiratory exchange ratio (A), total carbohydrate (CHO) oxidation (B), and ingested CHO oxidation (C) measured at successive 20-min intervals during 2 h of steady-state exercise. Values are means ± SE for 6 subjects. * Significantly different from Con; b significantly different from start of exercise for HGI, LGI, and Con; c significantly different from start of exercise for HGI and LGI (P < 0.05 for all symbols).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Total and ingested CHO oxidation during 2 h of steady-state exercise calculated from area under oxidation curves. Values are means ± SE for 6 subjects. Differences were not significant.

Subjects' RPE rose steadily throughout the steady-state exercise and did not differ among trials. The fluid deficits accumulated during exercise (estimated from changes in BM) were 1.2 ± 0.4, 1.3 ± 0.2, and 1.2 ± 0.1 kg for HGI, LGI, and Con trials, respectively. Performance during the time trial undertaken at the end of the 120 min of steady-state exercise did not differ among trials. Time to complete 300 kJ was 947 ± 23, 953 ± 36, and 970 ± 26 s, respectively, for HGI, LGI, and Con trials.

Although subjects were kept blind to their time trial performances until completion of the study, all were able to correctly identify the trial in which they performed best (HGI, n = 2; LGI, n = 3; Con, n = 1). When asked to choose the meal they would prefer to consume before an important competition, all subjects nominated the LGI (pasta) meal as being a familiar and preferred choice.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The major finding of this investigation was that when large amounts (~170 g) of CHO were ingested during prolonged (~2.5 h) cycling, there were minimal differences in the metabolic and performance responses to the choice of preevent meal. Preexercise meals and CHO intake during exercise were chosen to balance current guidelines for optimal sports nutrition practice with strategies that are realistic for competitive cycling. Specifically, subjects ate a meal 2 h before the exercise bout, which in the case of the CHO trials provided 2 g of CHO/kg BM. Just before exercise, a bolus of fluid was ingested to promote a rapid rate of nutrient delivery from the subsequent feedings (29). CHO was consumed throughout exercise (26) to provide a total intake of ~ 1 g/min; this is the maximal rate at which ingested CHO is oxidized during prolonged moderate-intensity exercise (19). This fluid and energy intake was achieved by consuming a 10 g/100 ml glucose drink at a rate of ~700 ml/h, a compromise between fluid intake recommendations (1) and rates typically achieved by athletes in competitive situations (28). Subjects tolerated their pre- and during-exercise feeding schedules; there were no reports of gastrointestinal discomfort.

This study was carried out in view of the historical controversy regarding CHO intake before prolonged submaximal exercise and the lack of application of preexercise feeding studies to the practices of competitive athletes. With regard to the former, there has been considerable focus on the metabolic disturbances caused by the rise in insulin concentration that accompanies preexercise CHO intake (12, 14, 17, 24). Even small elevations in plasma insulin suppress lipolysis during subsequent exercise (22); this effect is mostly likely to occur when CHO is consumed 30-60 min before exercise and insulin concentrations are raised at the onset of the bout. However, metabolic alterations have been observed when feedings were given 4 h before exercise, persisting even though glucose and insulin concentrations had normalized by the onset of exercise (9). Increased CHO oxidation during subsequent exercise may lead to a decrease in plasma glucose and/or an accelerated rate of muscle glycogen utilization. Indeed, Foster et al. (12) reported impaired cycle time to exhaustion at 80% of O_{2 max} when 75 g of glucose were fed 30 min before exercise, and publicity of this study caused widespread warnings to avoid CHO intake during the hour before endurance exercise. These fears have persisted despite evidence from at least a dozen subsequent studies that CHO intake during the hour before exercise enhances, or at least fails to affect, work capacity and performance of prolonged moderate-intensity exercise (for reviews see Refs. 8, 18). Furthermore, the intake of substantial amounts (~200 g) of CHO, which offset any increase in exercise CHO oxidation, enhances work capacity and performance (27, 31, 35).

The recent application of the GI to sports nutrition (33) has revived the prejudice about metabolic perturbations associated with preexercise CHO intake. Thomas and co-workers (33) reported enhanced work capacity when subjects consumed 1 g of CHO/kg BM from an LGI food (lentils), 1 h before cycling at 67% of O_{2 max}, compared with an equal amount of CHO eaten as an HGI food (potatoes). This benefit was attributed to lower glycemic and insulinemic responses to the LGI trial compared with the HGI meal, maintaining blood glucose concentrations during exercise, increasing FFA concentrations, and reducing exercise RER values.

Although this study (33) has led to widespread advice that endurance athletes should choose preexercise meals based on LGI CHO-rich foods and drinks (4), other investigations have failed to find that metabolic alterations translate into performance effects. A second study conducted by the same group (34) utilized the same prefeeding and exercise protocol; on this occasion, the test meals consisted of an LGI and an HGI powdered food, and an LGI and an HGI breakfast cereal. They reported a correlation between the meal GI and the subsequent depression of blood glucose and FFA concentrations during exercise; LGI meals were associated with higher glucose and FFA concentrations after 90 min than the HGI meals. Thus LGI meals appeared to provide a sustained source of CHO throughout the exercise bout and later recovery (34). However, there were no differences in exercise time to exhaustion among trials, and no correlation between exercise time and the GI of the meal (34). It should be noted that the measurement of performance used in both Thomas studies (33, 34) (cycling time to exhaustion at a fixed submaximal work rate) has a high (~25%) coefficient of variation (25), which increases the possibility of errors.

Febbraio and Stewart (11) found no differences in the performance of a time trial conducted after 2 h of cycling at 70% O_{2 max} when preexercise meals consisted of either an LGI CHO-rich food (lentils), HGI CHO-rich (potatoes), or a placebo (low-energy jelly). The meals were eaten 45 min before exercise and, in the case of the CHO meals, provided an intake of 1 g of CHO/kg BM. One point of difference in this study lies in the definition and measurement of "performance." The advantage of the exercise protocol employed by Febbraio and Stewart is that it provides a more sports-specific and reliable measurement of performance (coefficient of variation ~4%) (23), preceded by a period of steady-state exercise during which comparison of the metabolic responses to treatments can be made. Data from that study showed no differences in total CHO oxidation between the two CHO treatments and similar muscle glycogen utilization in all trials (11). The findings of that study are consistent with the view that glycemic differences at the onset of exercise are short lived and unimportant for the performance of most athletes.

Regardless of any effects of preexercise feedings on subsequent metabolism, the most effective and common strategy used by endurance athletes to promote CHO availability during exercise is to ingest CHO-rich drinks or foods during the event. CHO ingestion during exercise is important for endurance performance because it maintains euglycemia and high rates of CHO oxidation when endogenous CHO stores have become limited (6). This study has examined the effect of various preexercise CHO feedings and CHO intake during exercise on metabolism and subsequent performance. In agreement with other studies, we found that the ingestion of an HGI CHO-rich meal produced a greater glycemic and insulinemic response compared with an LGI CHO-rich meal (11, 32-34). However, the rise in insulin concentration after both CHO-rich meals caused a suppression of FFA concentrations. Insulin increases CHO oxidation (11) and suppresses lipolysis (22), even at very low concentrations; this appears to be an absolute rather than a dose-dependent effect.

Nevertheless, CHO feedings throughout exercise minimized any potential differences in either circulating blood metabolites or substrate oxidation during the bout. Blood glucose concentrations were maintained throughout exercise in all trials; there was no transient decline at the beginning of the bout as is typically seen with the preexercise intake of glucose (5, 10, 12, 16, 17, 24) or medium-GI and HGI CHO-rich foods (21, 22, 32, 33). High rates of CHO oxidation were sustained throughout the exercise bout; there was no difference at any time point among trials or over 2 h of exercise. The tracer-determined rates of oxidation of the ingested glucose were similar to those reported in other studies that have used a preexercise bolus feeding and serial feedings throughout exercise (for review, see Ref. 19). The LGI meal may have caused slower gastric emptying of the ingested drink. After 20 min of exercise, the rate of oxidation of the ingested glucose was lower in the LGI trial compared with the Con trial. However, this difference was short lived, and there were no differences among trials in the total oxidation of ingested glucose drink during 2 h of steady-state exercise. Irrespective of the choice of the preexercise meal, the ingested drink contributed ~60 g, or ~16%, of the total CHO oxidized. Given similar patterns of substrate utilization and availability among the three trials, it was not surprising to find similar performances during the timed rides that followed the steady-state exercise.

Wright et al. (35) examined the interaction of CHO intake before and during exercise. They found that, compared with no CHO intake at all, cycling time to exhaustion and total work output at 70% of O_{2 max} were improved by the ingestion of 5 g of CHO/kg BM 3 h before exercise or by intake of 2.6 g of CHO/kg BM in serial feedings during the trial. Enhancement of the performance measures was ~18 and 33% (P < 0.05) for the preexercise CHO and during exercise CHO intake, respectively. When undertaken together, the two strategies improved performance by ~45%. Whereas this suggests that the combination of CHO-intake strategies was superior to either of the feeding strategies alone, performances during the combined trial were not significantly different to the preexercise CHO trial or the during-exercise CHO trial (35). Similarly, in the present investigation, whereas there was no significant difference in ride time among trials, five of the six subjects achieved their best performance with a combination of CHO before exercise (HGI or LGI trial) and glucose intake during exercise, compared with the Con trial (CHO intake during exercise alone).

Finally, although subjects were able to correctly identify the trial in which they performed best, all reported that for an important competition they would choose to eat the preexercise meal that was most familiar. We conclude that the ingestion of CHO during prolonged moderate-intensity cycling according to current sports nutrition guidelines minimizes any differences in the metabolic and performance responses arising from the choice of preevent meal. Contrary to some advice that LGI CHO-rich foods are the preferred preexercise choice (4), endurance athletes are guided that when significant amounts of CHO are consumed during exercise they may choose their preexercise meal strategies in accordance with their personal preferences and previous experience.