Oxidation of Carbohydrate Feedings During Prolonged Exercise Current Thoughts, Guidelines and Directions for Future Research

Transcription

1 REVIEW ARTICLE Sports Med 2000 Jun; 29 (6): /00/ /$20.00/0 Adis International Limited. All rights reserved. Oxidation of Carbohydrate Feedings During Prolonged Exercise Current Thoughts, Guidelines and Directions for Future Research Asker E. Jeukendrup and Roy Jentjens Human Performance Laboratory, School of Sport and Exercise Sciences, University of Birmingham, Edgbaston, Birmingham, England Contents Abstract Methodological Considerations Radioactive Isotopes Stable Isotopes Feeding Strategies and Exogenous Carbohydrate (CHO) Oxidation Feeding Schedule Types of CHO Fructose Galactose Maltose Sucrose Glucose Polymers Maltodextrins Starch Summary Multiple Transportable CHOs Osmolality and Concentration Amount of CHO Factors Affecting Exogenous CHO Oxidation Exercise Intensity Muscle Glycogen Training Limitations of Exogenous CHO Oxidation Directions for Future Research Practical Implications, Guidelines and Conclusion Abstract Although it is known that carbohydrate (CHO) feedings during exercise improve endurance performance, the effects of different feeding strategies are less clear. Studies using (stable) isotope methodology have shown that not all carbohydrates are oxidised at similar rates and hence they may not be equally effective. Glucose, sucrose, maltose, maltodextrins and amylopectin are oxidised at high rates. Fructose, galactose and amylose have been shown to be oxidised at 25 to 50% lower rates. Combinations of multiple transportable CHO may increase the total CHO absorption and total exogenous CHO oxidation. Increasing the CHO

2 408 Jeukendrup & Jentjens intake up to 1.0 to 1.5 g/min will increase the oxidation up to about 1.0 to 1.1 g/min. However, a further increase of the intake will not further increase the oxidation rates. Training status does not affect exogenous CHO oxidation. The effects of fasting and muscle glycogen depletion are less clear. The most remarkable conclusion is probably that exogenous CHO oxidation rates do not exceed 1.0 to 1.1 g/min. There is convincing evidence that this limitation is not at the muscular level but most likely located in the intestine or the liver. Intestinal perfusion studies seem to suggest that the capacity to absorb glucose is only slightly in excess of the observed entrance of glucose into the blood and the rate of absorption may thus be a factor contributing to the limitation. However, the liver may play an additional important role, in that it provides glucose to the bloodstream at a rate of about 1 g/min by balancing the glucose from the gut and from glycogenolysis/gluconeogenesis. It is possible that when large amounts of glucose are ingested absorption is a limiting factor, and the liver will retain some glucose and thus act as a second limiting factor to exogenous CHO oxidation. The number of studies concluding that carbohydrate (CHO) feedings during exercise improve exercise capacity or exercise performance is so large that, from a scientific point of view, we can consider this relationship true. In the last few years, studies have accumulated to show that CHO feedings during exercise can positively affect performance when the exercise duration is about 45 minutes or longer. [1,2] The mechanism by which these CHO feedings exert their effect is believed to be a maintenance of blood glucose and increased rates of CHO oxidation during exercise. [2] It has also been shown that CHO feedings during exercise spare liver glycogen. [3-5] However, whether CHO feedings spare muscle glycogen is still controversial, as some studies reported glycogen sparing [6,7] whereas others did not. [2,8] This debate has recently been reviewed by Tsintzas and Williams. [9] Several studies have also addressed the questions of which CHO was most effective, what the most effective feeding schedule was and the optimal amount of CHO to be ingested. Additional studies have looked at factors that can possibly influence the oxidation of ingested CHO, such as muscle glycogen levels, diet, and exercise intensity. More recently, studies have attempted to detect the factors that limit the maximal rates of exogenous CHO oxidation. The purpose of this review is not to review the effects of CHO on exercise performance per se,but to summarise the factors that determine the efficacy (i.e. oxidation) of ingested CHO. With the conclusions from this overview, guidelines will be formulated for the use of CHO supplements during exercise. Finally, some of the remaining questions and directions for future research will be discussed. 1. Methodological Considerations TheoxidationofingestedCHOcanbemeasured by using isotope techniques. Costill et al. [10] were probably the first to study the oxidation of ingested CHO. They labeled the CHO in a drink with a radioactive tracer ([U- 14 C]glucose) and reported that only a small amount of an ingested CHO load was oxidised during exercise. As a result, they concluded that CHO feedings were of limited importance for muscle metabolism. However, this result was probably the result of methodological problems, since many studies in the following years have shown significant contributions of ingested CHO to energy expenditure during exercise. Most studies today use stable isotopes for the measurement of exogenous CHO oxidation, since this does not provoke any health hazards in contrast to the potential negative effects of radioactive isotopes. The advantages and

3 Oxidation of Carbohydrate Feedings During Exercise 409 disadvantages of these techniques will be discussed in sections 1.1 and Radioactive Isotopes The oldest method to trace ingested CHO is to add a [U- 14 C]glucose tracer to a CHO beverage and measure 14 C in expired gases using a scintillation counter. The advantage of this technique is that it is relatively inexpensive compared with the use of stable isotopes. In addition, shifts in background enrichments which may occur when using stable isotopes (see section 1.2) are not a problem, because the background level of 14 C is negligible. An obvious disadvantage of this technique is the fact that it exposes the volunteer to radioactivity. Although the radiation dose given is usually low (<40 uci/l is consumed), and is calculated to correspondto0.02to0.03rem,200to250timeslower than the permissible dose, the actual risks may often be underestimated. [11] Glucose is not only used for oxidation, but is also a substrate for other metabolic pathways, including pathways that result in the formation of DNA. Incorporation of radioactivity in a DNA molecule is of course dangerous because it may damage genetic material. It is therefore advisable to use stable isotopes rather than radioactive isotopes to study metabolism. One potential problem with using isotopes (radioactive or stable) is that part of the CO 2 (including 14 CO 2 or 13 CO 2 ) may not appear in the expired gases because it is temporarily trapped in the bicarbonate pool. CO 2 +H 2 O à H 2 CO 3 à HCO 3 +H + (Eq. 1) Thisisaverylargeandonlyslowlyexchanging pool, in which CO 2, arising from various decarboxylation reactions, is retained. In resting conditions, it may take hours before there is an equilibrium between 14 CO 2 and H 14 CO 3 (or 13 CO 2 and H 13 CO 3 ). However, during exercise the turnover of this pool increases severalfold and, especially at high absolute workloads, equilibrium may be reached within 60 minutes. It has been reported that recovery of 13 CO 2 approached 100% after 60 minutes of exercise at 60 to 70% maximal oxygen uptake (V. O 2max ). [10,12,13] In many experimental conditions, the entrapment of 14 CO 2 or 13 CO 2 in the bicarbonate pool may cause a marked underestimation of the true exogenous CHO oxidation, especially during the first hour of exercise. There are a few ways around this problem. One way is to prime the bicarbonate pool with H 14 CO 3 or H 13 CO 3. This would bring the bicarbonate pool into equilibrium within the first 15 minutes of exercise. [5,8] A second way is to avoid calculating exogenous CHO oxidation rates in the first hour. [14] Finally, it is possible to use an acetate correction factor as suggested recently. [15] In addition to the temporary label loss in the bicarbonate pool, it has also been reported that, in studies using a 13 C-tracer for studying fatty acid metabolism, part of the tracer may be trapped in exchange reactions with the tricarboxylic acid (TCA)-cycle. [15,16] For example, some 13 C-carbons may be incorporated into the glutamate/glutamine pool via α-ketoglutarate (α-kg), or into phosphoenolpyruvate (PEP) via oxaloacetate (OAA). [16] This label fixation results in a decreased recovery of label in the expired gases and, in order to correct for this loss, the acetate correction factor has been proposed. [15] This correction is based on the assumption that acetate has immediate access to the TCA-cycle and is instantly oxidised. The percentage of label ( 13 Cor 14 C) not recovered in expired CO 2 represents the amount of CO 2 trapped in exchange reactions with TCA-cycle intermediates (TCAI) and the bicarbonate pool. The label loss is dependent on the metabolic rate. At high oxygen uptakes (>35 ml/kg/min) less label is trapped and recovery of the 1-14 C-acetate label was foundtobe85to90%. [15] Similar results were obtained by Schrauwen et al. [16] when [U- 13 C]palmitate was used. This implies that studies performed at low absolute exercise intensities may have underestimated exogenous CHO oxidation rates. 1.2 Stable Isotopes Studies in which stable isotope methodology was used to measure exogenous CHO oxidation have used 13 C-enriched substrates. Some of these studies have used naturally enriched CHO (derived

4 410 Jeukendrup & Jentjens from C4 plants such as corn and cane sugar). These plants have a naturally high abundance of 13 C. When ingesting these CHOs during exercise, breath 13 CO 2 will become enriched and, together with a measure of the total CO 2 production rate, exogenous CHO oxidation rates can be quantified. In addition to the problems described above, there is another complication with this technique: shifts in substrate utilisation may result in a change in background enrichment. [17,18] Because CHO is usually more 13 C-enriched than fat, glycogen stores may display higher 13 C-enrichments than endogenous fat stores. Any change in shift in endogenous substrate utilisation can therefore cause a change in the background 13 C-enrichment independent of ingested CHO. These changes occur for instance in the transition from rest to exercise, and typically an increase in 13 CO 2 in the expired gases is observed. The magnitude of the error depends on the 13 C-enrichment of the ingested CHO relative to the 13 C-enrichment of endogenous glycogen stores. It has been shown that individuals with a diet in which most CHOs are derived from C4 plants (a typical northern American or Canadian diet) have higher 13 C-enrichments in their muscle glycogen stores compared with Europeans, whose diet is typically derived from C3 plants such as potato and beet sugar. In a comparative study at 60% V. O 2max at Ball State University (Indiana, USA) and Maastricht University (The Netherlands), we have observed that in northern America, shifts in background enrichment may be 3 to 5 times higher than in Europe (unpublished data). Several investigators have therefore instructed their study participants not to consume products with a high natural 13 C-abundance, or have reduced the error by artificially increasing the 13 C-enrichment of the CHO ingested during the experiment (typically by adding [U- 13 C]glucose to a CHO beverage). By adding a tracer to the CHO, the shift in background remains the same but the relative error is reduced. Another way around the problem is to perform control trials with an identical protocol but with ingestion of CHO with a low natural abundance. The background 13 C-enrichments can then be used to correct the calculated exogenous CHO oxidation. Exogenous CHO oxidation = V. CO 2 (ECO 2 Ebkg)/(Eing Ebkg) 1/k (Eq. 2) where V. CO 2 is the total CO 2 production rate, ECO 2 is the 13 C-enrichment of CO 2,Eingisthe 13 C- enrichment of the ingested CHO, Ebkg is the background enrichment determined in a separate experiment with the same conditions, and k is the amount of CO 2 that will arise from the oxidation of 1g of glucose (0.7466L CO 2 /g glucose). It is possible to obtain accurate and reliable measures of exogenous CHO oxidation using (radioactive or stable) isotopes. However, as was just discussed, there are several errors that can be made andhavebeenmadeinthepast.thisisimportant when interpreting results, especially from some of the earlier studies. The absolute values reported in several trials may be overestimated in studies using CHO with a naturally high 13 C-abundance because no corrections were made for background enrichment. Other studies may have underestimated exogenous CHO oxidation because no correction was made for label loss or label fixation. We would like the reader to keep this in mind when interpreting the results of various studies. Here, we will present the data of different studies as presented in the original papers. We have not tried to correct for the possible methodological errors because there were too many unknown variables (e.g. diet, background enrichments) and often papers did not report sufficient information (e.g. enrichment data) to allow these corrections to be made. Nevertheless, in most cases the error will be small (5 to 10%) and correction would not have altered the conclusions of these papers since typically 2 or 3 trials are compared in the same experimental conditions. 2. Feeding Strategies and Exogenous Carbohydrate (CHO) Oxidation 2.1 Feeding Schedule The typical pattern of exogenous glucose oxidation rates is shown in figure 1. The first appearance

5 Oxidation of Carbohydrate Feedings During Exercise 411 of label from ingested CHO can already be observed in the first 5 minutes (unpublished observations). During the first 75 to 90 minutes of exercise, exogenous CHO oxidation will continue to rise as more and more CHO will be emptied from the stomach and absorbed in the intestine. After 75 to 90 minutes a leveling-off will occur and the exogenous CHO oxidation rate will reach its maximum value and will not further increase. The timing of CHO feedings seemed to have very little effect on the slope of this curve or the plateau value. In several studies [19-23] the oxidation of a single glucose load (100g) given at the onset of exercise (90 to 120 minutes) was investigated. They all reported a very similar oxidation pattern for ingested glucose; an increase in oxidation rates during the first 75 to 90 minutes and a plateau thereafter. Maximal exogenous CHO oxidation rates in these studies varied between 0.48 and 0.65 g/min. These rates are similar to those observed when ingesting similar amountsofglucose(90to100gin90to120minutes) as repetitive feedings during exercise. [24-28] In a study by Krzentowski et al., [20] volunteers walked at a 10% grade (45% V. O 2max ) for 4 hours. They ingested 100g of glucose after 15 or 120 minutes. Exogenous CHO oxidation rates followed an identical pattern from the time of ingestion until 2 hours later. The amount of ingested glucose oxidised was similar in the 2 hours following ingestion (55g when CHO was ingested after 15 minutes and 54g when ingested after 120 minutes). This study showed that the time of ingestion has no effect on exogenous CHO oxidation. Often repetitive feeding schedules are adopted because it has been shown that this accelerates the rate of gastric emptying and hence the delivery of CHO to the intestine. [29,30] However, since gastric emptying does not usually limit exogenous CHO oxidation, [27,31] the feeding schedule may have little effect on the maximum oxidation rates or the time to reach these high rates of oxidation. Thus, although there are no studies available that have directly studied the effect of different feeding schedules on the rate of exogenous CHO oxidation, the literature seems to suggest that the feeding schedule has very little Exogenous CHO oxidation (g/min) impact on the maximal exogenous CHO oxidation rates or the time to reach this maximum. However, the feeding schedule should be such that high exogenous CHO oxidation rates are achieved as soon as possible after the onset of exercise and the amount of CHO ingested should be sufficient to maintain high rates of exogenous CHO oxidation. McConell et al. [32] compared the effects of CHO ingestion throughout exercise with ingestion of an equal amount of CHO late in exercise. In this study, performance was improved relative to the control trial only when CHO was ingested throughout exercise. CHO ingestion late in exercise did not improve performance despite increases in plasma glucose and insulin levels. 2.2 Types of CHO HI-GLU 0.2 LO-GLU Time (min) Fig. 1. Typical pattern of exogenous carbohydrate (CHO) oxidation during exercise when beverages are consumed at the onset of exercise and at regular intervals thereafter. HI-GLU = high glucose ingestion; LO-GLU = low glucose ingestion. In figure 2, different types of dietary CHO are depicted. Different types of CHO may have different properties. Differences in osmolality and structure have effects on taste, digestion, absorption, the release of various hormones, and the availability of glucose for oxidation in the muscle. A number of studies have compared the oxidation rates of various types of ingested CHO with the oxidation of glucose during exercise. [26,27,31,33-38] The results will be discussed in the following sections Fructose There has been considerable interest in fructose for a variety of reasons. [23,39,40] The first reason is

6 412 Jeukendrup & Jentjens C OH Glucose Fructose Galactose OH C C C O OH C C OH OH Maltose Sucrose Lactose Amylopectin starch Maltodextrin Amylose starch Fig. 2. Overview of different carbohydrates and their structure. There are 3 monosaccharides (glucose, fructose and galactose) and 3 disaccharides (maltose, sucrose and lactose). Glucose polymers (maltodextrins) and starch consist of a series of coupled glucose molecules. that adding fructose will generally improve the palatability of a drink. Secondly, fructose will cause a 20 to 30% smaller increase in plasma insulin levels compared with glucose, [41] and hence it will reduce lipolysis to a smaller extent. Fructose has also been used as a pre-exercise feeding to prevent exerciseinduced rebound hypoglycaemia. [23,39,40] Massicotte and colleagues [26,33] studied the oxidation of fructose compared with an isoenergetic glucose solution and found 25% lower oxidation rates for fructose. Jandrain et al. [42] studied exogenous CHO oxidation rates in 10 healthy but untrained volunteers during 3 hours of exercise at 45% V. O 2max while ingesting 150g glucose or fructose. The peak oxidation rates for the ingested glucose were 0.67 g/min and fructose oxidation peaked at 0.50 g/min (25% lower). Similar findings were reported by others. [23,34,43,44] The lower oxidation rates of fructose are probably due to a lower rate of absorption and the fact that fructose has to be converted into glucose in the liver before it can be metabolised. The latter is usually a relatively slow process. Interestingly, during fasting when gluconeogenic pathways are activated, similar rates of oxidation were found for glucose and fructose. [25,34] Galactose Only one study has investigated the oxidation rates of ingested galactose during exercise. Leijssen et al. [35] fed 8 volunteers, who exercised for 2 hours at 70% V. O 2max, 155g of galactose or glucose and calculated the oxidation rates of the exogenous CHO. While glucose was oxidised at a rate of 0.85 g/min during the last hour, galactose oxidation was only half of that (0.41 g/min). It was suggested that the absorption or the conversion into glucose in the liver was limiting. Galactose on its own therefore seemed an inappropriate source of CHO for sports drinks Maltose Hawley et al. [36] investigated the oxidation of maltose and glucose during 90 minutes of exercise at 70% V. O 2max. Trained volunteers ingested 180g of glucose or maltose during exercise and exogenous CHO oxidation was measured using radioactive isotopes. High peak oxidation rates were reached at the end of exercise and equaled 0.9 g/min for glucose and 1.0 g/min for maltose. These differences were not statistically significant and it was concluded that maltose and glucose are oxidised at

7 Oxidation of Carbohydrate Feedings During Exercise 413 similar rates. In addition, these authors found no differences in the absorption rates of these CHOs Sucrose Few studies have investigated the oxidation of ingested sucrose. In a study by Moodley et al., [27] volunteers ingested 90g of sucrose during 90 minutes of exercise at 70% V. O 2max. Sucrose oxidation rates peaked at approximately 0.4 g/min. Although these rates may seem quite low, similar oxidation rates were reported for glucose and the low values may therefore be a result of the methodology used in that study. Wagenmakers et al. [37] gave their study participants an 8% sucrose solution during 2 hours of cycling exercise at 65% V. O 2max. The total amount of sucrose ingested during the 2 hours was 145g, and it was estimated that 81g was oxidised. The peak oxidation rate was 0.87 g/min, a value similar to that observed after glucose ingestion in other studies. [8,14,25-28,36,45] It can therefore be concluded that sucrose can be oxidised at similar rates as glucose and the efficacy of these 2 CHOs may be similar Glucose Polymers Maltodextrins Because of their neutral taste and their relatively low osmotic value, maltodextrins have been used by many manufacturers of sports drinks to increase the CHO content of these beverages. In a study by Rehrer et al., [31] a 17% maltodextrin solution was compared with a 17% glucose solution. The total amount of CHO that was ingested during 80 minutes of exercise at 70% V. O 2max was 220g. Oral CHO oxidation was measured and was found to be similar for the glucose and the maltodextrin drink (42 and 39g for glucose and maltodextrin, respectively). A peak oxidation rate of 0.78 g/min was reported for glucose and 0.75 g/min for maltodextrins. These results indicate that there is no difference in the oxidation of maltodextrins and glucose. In addition, it was found that the rates of gastric emptying and thus the rate of delivery of CHO to the intestine was similar between glucose and the glucose polymer. These results also imply that the digestion (hydrolysis of the bonds between glucose molecules of a glucose polymer) is not a rate-limiting step for exogenous CHO oxidation. Table I. Amylose and amylopectin content of various plant starches Plant starches Amylose (%) Amylopectin (%) Maize Potato Rice Tapioca Wheat Wagenmakers et al. [37] found similar results when feeding volunteers maltodextrin solutions ranging from 4 to 16%. Increasing rates of CHO ingestion seemed to increase oral CHO oxidation up to a rate of 1.0 to 1.1 g/min. Ingestion of more than 1.2 g/min had very little or no additional effect on the oxidation rates. [37] However, these high rates of ingestion did result in high oral CHO oxidation rates (0.53 to 1.07 g/min) that were similar to the rates observed with glucose ingestion in other studies Starch There are 2 major types of starch: amylopectin and amylose. Amylopectin is a highly branched molecule, whereas amylose is a long straight chain of glucose molecules (fig. 2) twisted into a helical coil. Branches in starch are created by 1,6 bonds between glucose units, whereas 1,4 glucosidic bonds will result in a straight chain of glucose units. Starches with a relatively large amount of amylopectin are rapidly digested and absorbed, whereas those with a high amylose content will have a slow rate of hydrolysis. Starches make up approximately 50% of our total daily CHO intake and most naturally occurring starches are a mixture of amylose and amylopectin (see table I). One study [38] compared the rate of gastric emptying and the oxidation rate of an insoluble starch consisting of 23% amylose and 77% amylopectin with a soluble starch consisting of 100% amylopectin. Volunteers ingested 316g during 2.5 hours of cycling exercise at 68% V. O 2max. The amount of CHO delivered to the intestine seemed somewhat lower in the case of the insoluble starch, but this difference did not reach statistical significance. However, the insoluble starch was oxidised at a lower rate (75g of insoluble starch compared with 126g of soluble starch). Peak oxidation rates were 1.1 and 0.8 g/min for the soluble

8 414 Jeukendrup & Jentjens Oral CHO oxidation rate (g/min) Glucose Fructose Galactose Sucrose Maltose MD Starch CHO ingestion rate (g/min) Fig. 3. Peak oxidation rates of oral carbohydrates (CHOs) are depicted against the CHO ingestion rate of different types of CHO. Fructose and galactose appear to be oxidised at relatively low rates whereas glucose, sucrose, maltose, maltodextrins and soluble starch seem to be oxidised at relatively high rates. The horizontal line depicts the absolute maximum for oral CHO oxidation. The dotted line represents the line of identity, where CHO ingestion equals CHO oxidation. starch and the insoluble starch, respectively, while the insoluble starch seemed to cause some gastrointestinal discomfort. [38] The oxidation of amylose only was not measured but can be assumed to be very low. Although one study reported a very high rate of oxidation for insoluble starch, [46] this has been shown to be due to a methodological error. [38] In conclusion, amylopectin is oxidised at higher rates than amylose and is therefore a more appropriate energy source in CHO beverages for athletes. Furthermore, insoluble starch may provoke gastrointestinal symptoms. [38] Summary The results of various studies are summarised in figure 3. This figure shows the peak oxidation rates, which may depend on a variety of factors including the exercise intensity, the amount of CHO ingested, and the timing of these feedings. Fructose and galactose appear to be oxidised at relatively low rates, whereas glucose, sucrose, maltose, maltodextrins and soluble starch seem to be oxidised at relatively high rates. Maximal oral CHO oxidation seems to be around 1 g/min. The horizontal line depicts the absolute maximum just below 1.1 g/min. The dotted line represents the line of identity, where CHO ingestion equals CHO oxidation. From this graph it can be concluded that oral CHO oxidation may be optimal at rates of ingestion around 1.0 to 1.5 g/min. This implies that athletes should ensure a CHO intake of about 60 to 70g per hour for optimal CHO delivery. Adopting an ingestion rate of 60 to 70 g/h will optimise exogenous CHO oxidation. 2.3 Multiple Transportable CHOs A study by Shi and colleagues [47] suggested that the inclusion of 2 or 3 CHOs (glucose, fructose and sucrose) in a drink may increase water and CHO absorption despite increased osmolality. This effect was attributed to the separate transport mechanisms across the intestinal wall for glucose, fructose and sucrose. [47] Interestingly, fructose absorption from sucrose is also more rapid than the absorption of an

9 Oxidation of Carbohydrate Feedings During Exercise 415 equimolar amount of fructose. In an elegant study, Adopo et al. [44] fed 6 volunteers CHO at the onset of 2 hours of exercise at 61% V. O 2max. The CHO feedings were 50g of glucose, 50g of fructose, 100g of glucose, 100g of fructose or 50g of glucose plus 50g of fructose. It was found that adding fructose to a glucose solution increases the oral CHO oxidation by 21% compared with an iso-energetic glucose solution (fig. 3). The oxidation rate of 50g glucose plus 50g fructose in a combined drink was higher than the oxidation rate of either 100g glucose or 100g fructose. However, amounts ingested were relatively small and it remains to be established whether combined ingestion of glucose and fructose can increase exogenous CHO oxidation more than the ingestion of large amounts of a single CHO. Whether addition of galactose to a glucose drink can increase total exogenous CHO oxidation in a similar way to glucose and fructose needs to be determined. These data suggest that it might be useful to include multiple types of CHO in CHO drinks for athletes. More studies are needed to identify optimal combinations of different CHOs. 2.4 Osmolality and Concentration Gastric emptying and absorption may depend on the concentration and osmolality and hence the type and amount of CHO, and the volume of the ingested beverage. Recent studies seem to suggest that CHO content is a more important determinant of gastric emptying than osmolality. [48] Therefore, the CHO type may have little or no effect on the rate of gastric emptying. [49] It has become clear that the CHO type and osmolality of a solution can influence intestinal absorption of fluid and CHO. Relatively large amounts of glucose in the form of glucose polymers introduced to the gastrointestinal tract without changing the osmotic load can increase the glucose delivery and induce greater water absorption. [50] Jandrain et al. [19] investigated the oxidation of a 50g glucose load dissolved in either 200, 400 or 600ml of water. Although both the concentration and osmolality were different in these drinks, no differences were observed in exogenous CHO oxidation during 4 hours of exercise at 45% V. O 2max.This study suggests that the total amount of CHO seems to be a more important determinant of exogenous CHO oxidation than osmolality or CHO concentration. 2.5 Amount of CHO The amount of CHO that needs to be ingested in order to obtain optimal performance is important from a practical point of view. The optimal amount is likely to be the amount of CHO resulting in maximal exogenous CHO oxidation rates. Pallikarakis et al. [51] found that doubling the amount of CHO ingestedfrom200to400gduring285minutesof exercise at 45% V. O 2max increased exogenous CHO oxidation. However, exogenous CHO oxidation rates did not double and the percentage of the CHO ingested that was oxidised was slightly lower (59.5 and 56.8%, respectively). Here we will refer to this phenomenon as a lower oxidation efficiency with thelargerdoseofcho. Oxidation efficiency = exogenous CHO oxidation rate/ingestion rate 100% (Eq. 3) Rehrer et al. [31] studied the oxidation of different amounts of CHO ingested during 80 minutes of cycling exercise at 70% V. O 2max. In a randomised cross-over design, volunteers received a 4.5% glucose solution (a total of 58g glucose during 80 minutes of exercise) or a 17% glucose solution (220g during 80 minutes of exercise). Exogenous CHO oxidation was measured and these were slightly higher with the larger CHO dose (42 and 32g in 80 minutes, respectively). Thus, even though the amount of CHO ingested was increased almost 4-fold, the oxidation rates were barely affected. The oxidation efficiency was much lower with the large amount of CHO (19% for the 17% glucose solution versus 55% for the 4.5% glucose solution). Ingestion of a 17% maltodextrin solution lead to the same conclusion (i.e. there was a lower oxidation efficiency with the more concentrated solution). In a study by Wagenmakers et al., [37] participants exercisedfor120minutesat65%v. O 2max on 5 occasions and received 4 doses of maltodextrin ranging

10 416 Jeukendrup & Jentjens from 72 to 289g. Calculated average ingestion rates were 0.6, 1.2, 1.8 and 2.4 g/min. Although oxidation rates increased with increasing intake, exogenous CHO oxidation seemed to level off after an intake of 1.2 g/min. Oxidation rates were 0.53, 0.86, 1.00 and 1.07 g/min, respectively. Also in this study, the oxidation efficiency decreased with increasing intake (72, 52, 39 and 32%, respectively). More recently, Jeukendrup et al. [5] investigated the oxidation rates of even larger CHO intakes on exogenous CHO oxidation. In this study, well trained volunteers exercised at a relatively low exercise intensity of 50% V. O 2max for 120 minutes while ingesting 70 or 360g of glucose. With the low dose of glucose (average ingestion rate of 0.58 g/min) exogenous CHO oxidation rates averaged 0.34 g/min, while with the high dose (average ingestion rate 3.00 g/min) these rates increased up to 0.94 g/min. This study also demonstrated a decreased CHO oxidation efficiency with increasing ingestion rates (59 vs 31%). It is interesting to note that although ingestion rates increased up to 2.4 to 3.0 g/min, [5,37] in none of these studies did CHO oxidation rates exceed 1.1 g/min. The results of all studies currently available in the literature were used to construct figure 3. Although this graph needs to be interpreted with caution (it includes studies at different exercise intensities, different feeding schedules, different volunteer populations, etc.), it must be concluded that the maximal rate at which ingested CHO can be oxidised is 1.0 to 1.1 g/min. Increasing the CHO intake during exercise may increase oxidation rates until the intake exceeds 1.0 to 1.2 g/min. Clearly, the rate of oxidation of ingested CHO is limited. However, the factors limiting exogenous CHO oxidation are still largely unknown. Possible mechanisms will be discussedinsection3. 3. Factors Affecting Exogenous CHO Oxidation 3.1 Exercise Intensity With increasing exercise intensity, the exercising muscle becomes more and more dependent on CHO as a source of energy. Both an increased muscle glycogenolysis and increased plasma glucose oxidation will contribute to the increased energy demands. [52] It is therefore reasonable to suspect that exogenous CHO oxidation might increase with increasing exercise intensities. Indeed, an early study by Pirnay et al. [53] reported lower exogenous CHO oxidation rates at low exercise intensities compared with moderate intensities, but exogenous CHO oxidation tended to level off between 51 and 64% V. O 2max. In this study, participants exercised for 90 minutes on a treadmill on 4 different occasions at different percentages of their maximal aerobic capacity. They ingested 100g of glucose during exercise. The average oxidation rates of the ingested glucose were 0.18, 0.36, 0.46 and 0.49 g/min at 22, 39, 51, and 64% V. O 2max, respectively. The exogenous CHO oxidation rates did not further increase when the exercise intensity was increased from 51 to 64% V. O 2max. Recently, the same group of researchers found an almost similar relationship between the exogenous CHO oxidation rate and the power output on a cycle ergometer. [54] The oxidation rate of the ingested CHO increased with increasing metabolism for intensities below 60% V. O 2max. However, when the exercise intensity was increased from 60 to 75% V. O 2max the oxidation rate leveled off or even decreased (0.51 and 0.42 g/min, respectively). One possible explanation for the reduced exogenous oxidation rate during high exercise intensities (>70 to 75% V. O 2max ) might be the limitation of intestinal digestion and/or absorption, although to our knowledge such a limitation has not been shown at exercise intensities below 80% V. O 2max. Massicotte et al. [28] examined a group of individuals with a wide variety of fitness levels during exercise at 60% of their individual V. O 2max. Although volunteers exercised at the same relative workload (60% V. O 2max ), there were large differences in the metabolic rate (absolute workload). In agreement with the findings of Pirnay et al., [53,54] a linear relationship between the metabolic rate and the oxidation rate of 100g ingested CHO was found.

11 Oxidation of Carbohydrate Feedings During Exercise 417 However, it could be argued that these findings are an artifact caused by the stable isotopic methods used, rather than a physiological phenomenon. As discussed in section 1, some label may be lost in exchange reactions with the TCA-cycle. It was also shown that at low metabolic rates recovery of the label was only 60 to 70%, whereas at high work rates recovery of the label can be 90% or more. [13] Because no correction was made for label loss in the studies cited above, the calculated exogenous CHO oxidation rates could have been underestimated, especially at lower metabolic rates. We therefore corrected the values for label loss according to Sidossis et al. [13] However, although the differences were less pronounced after correction, they were still present. Van Loon et al. [55] did not observe differences in exogenous CHO oxidation rates when trained cyclists exercised at 38 or 55% V. O 2max. It is therefore possible that lower exogenous CHO oxidation rates are only observed at very low exercise intensities when the reliance on CHO as an energy source is minimal. In this situation, part of the ingested CHO may be directed towards non-oxidative glucose disposal (storage in the liver or muscle) rather than towards oxidation. Studies with CHO ingestion during intermittent exercise have suggested that glycogen can be resynthesised during low intensity exercise. [56] It seems fair to conclude that at exercise intensities below 50 to 60% V. O 2max, exogenous CHO oxidation will increase with increasing total CHO oxidation rates, whereas above approximately 50 to 60% V. O 2max, oxidation rates will not usually increase further. 3.2 Muscle Glycogen Although determinants of exogenous CHO oxidation have been intensively investigated for almost 30 years, the effect of pre-exercise glycogen levels on exogenous CHO oxidation during exercise are still largely unknown and studies have produced different results. In a study conducted by Ravussin et al., [57] the oxidation rate of exogenous glucose was studied in individuals with normal and low glycogen levels. The 2 groups were observed for 2 hours at 40% V. O 2max on a cycle ergometer, 1 hour after ingestion of 100g of glucose. The oxidation rates of the ingested CHOs were similar: 41g in the group with normal glycogen availability and 38g in the group with reduced glycogen availability. However, the study had no cross-over design, which may have influenced the results. Although the absolute rates of exogenous CHO were not different between groups, due to the 20% higher energy expenditure observed in the group of glycogendepleted individuals, exogenous CHO oxidation provided only 16% of the energy yield versus 20% in the group with normal glycogen levels. Thus, the lower glycogen level was associated with a decreased contribution of exogenous CHO oxidation to energy expenditure during moderate intensity exercise. More recently, Jeukendrup et al. [45] manipulated pre-exercise glycogen levels by glycogen lowering exercise in combination with CHO restriction (LG trial) or rest in combination with CHO loading (HG trial). In a randomised cross-over design, volunteers received an average of 127g glucose during 120 minutes of exercise at 57% V. O 2max.Incontrast to the conclusion of Ravussin et al., [57] it was found that exogenous glucose oxidation was 28% lower in the LG trial compared with the HG trial: 36g of glucose was oxidised during 60 to 120 minutes of exercise during LG, whereas 50g was oxidised with HG. Péronnet et al. [58] studied the effect of endogenous CHO availability, after high and low CHO diets, on the oxidation of exogenous CHOs during 120 minutes of exercise at 64% V. O 2max.Volunteers relied more on exogenous CHO oxidation after the low CHO diet, when glycogen availability was presumably low, than after the high CHO diet, when glycogen availability was presumably high. Between 40 and 80 minutes of the exercise period, exogenous CHO oxidation was significantly higher after the low CHO diet compared with the high CHO diet (0.63 vs 0.52 g/min, respectively). These results are inconsistent with the results of Ravussin et al. [57] and Jeukendrup et al., [45] and are likely attributed to differences in experimental conditions of exercise and the amounts of CHO ingestion.

12 418 Jeukendrup & Jentjens Because of the higher relative workload in the study of Péronnet et al. [58] (64 vs 40 and 57% V. O 2max in the studies by Ravussin et al. [57] and Jeukendrup et al., [45] respectively) and the larger amount of glucose ingested (200 vs 100 and 127g in the studies by Ravussin et al. [57] and Jeukendrup et al., [45] respectively) volunteers relied more on CHO oxidation and less on fat oxidation after both diets. The increased reliance on CHO oxidation at this higher exercise intensity, when glycogen levels are reduced, might explain why exogenous CHO oxidation was higher. Another explanation could be that the extent to which glycogen levels were reduced was responsible for the different findings between the studies. Although none of the above studies measured glycogen levels, the glycogen depletion protocol used in the study by Jeukendrup et al. [45] has previously been shown to result in very low muscle glycogen levels (<140 mmol/kg dry weight), [59] and to lead to low plasma insulin levels and high plasma free fatty acids. The 2 to 3 times higher plasma free fatty acid level and the lower plasma insulin level when glycogen levels were low [45,57] could have reduced plasma glucose uptake and oxidation. [60] Péronnet et al. [58] found a smaller difference in free fatty acid levels between their experimental trials, whereas insulin levels were not different. This was possibly due to the moderate glycogen depletion regimen applied in their study, which might therefore explain why exogenous CHO oxidation did not decrease when glycogen availability was low. The effect of muscle glycogen on exogenous CHO oxidation per se is unknown at present. Studies have attempted to manipulate muscle glycogen stores by altering the dietary CHO intake and employing exercise programmes but, by doing so, other variables (i.e hormonal changes, high free fatty acid levels) have been changed as well and these changes may have been responsible for the variable results in different studies. More studies are required to elucidate the role of muscle glycogen on the oxidation rate of ingested CHO. 3.3 Training Endurance training has a marked effect on substrate utilisation and generally results in a shift from CHO towards fat metabolism. There is a decreased reliance on CHO metabolism after training at the same absolute workload. [61-64] However, some controversy still exists regarding whether the reliance on CHO as a fuel is also decreased at the same relative exercise intensity. Several studies suggest that even though the exercise after training is performed at the same relative intensity (and thus a higher absolute intensity), there is a decreased reliance on blood glucose and muscle glycogen. [63-65] However, some studies did not find a change in glucose uptake after training when compared at the same relative exercise intensity, [61,62] although plasma glucose oxidation was decreased. [62] Training induces several adaptations at the muscular level including an increased GLUT-4 content, [66] increased insulin action [67] and an increased capillary bed. All these adaptations would favour glucose uptake and could possibly alter the handling of blood glucose and thus of exogenous glucose. A few studies have investigated the effects of training (or training status) on exogenous CHO oxidation rates. [24,55,68,69] In an early study by Krzentowski et al., [68] volunteers trained for 6 weeks and substrate utilisation was measured at the same absolute exercise intensity before and after the training programme. The authors concluded that exogenous CHO oxidation was increased by 17% after training. However, the results seem difficult to interpret. Firstly, the V. O 2max of the participants was increased by an unphysiological amount (29%). Secondly, in contrast to the literature and despite the improved aerobic capacity after training, no difference in total CHO and fat oxidation was observed. More recently, van Loon et al. [55] reported that the contribution of CHO to energy expenditure was lower in well trained cyclists compared with healthy untrained controls at the same absolute intensity. The reduction in CHO oxidation was due to a reduction in muscle glycogen oxidation (0.10 and 0.75 g/min) and endogenous glucose production (0.20 and 0.13 g/min), respectively (fig. 4). However, de-

13 Oxidation of Carbohydrate Feedings During Exercise 419 Ra glucose (g/min) Glucose from liver Glucose from feedings 0 Fast LO-GLU HI-GLU Fig. 4. Glucose delivery to the blood from the liver and gastrointestinal tract (feedings) during exercise. During a fast, no glucose feedings were provided and all glucose appearing in the blood stream was derived from the liver. When a small amount of glucose was provided (LO-GLU) the total delivery of carbohydrate (CHO) increased but the contribution of liver glucose declined. When large amounts of CHO were ingested (HI-GLU), the total delivery of CHO was further increased. Liver glucose output was negligible and all glucose was derived from the feedings. Ra = rate of appearance (adapted from Jeukendrup et al., [70] with permission). spite these differences in substrate utilisation, exogenous glucose oxidation rates were unaffected (0.7 g/min in trained and untrained cyclists). Burelle et al. [24] also compared exogenous CHO oxidation in trained and sedentary individuals during exercise at the same absolute workload. Volunteers cycled for 90 minutes at 140W and received 100g 13 C-enriched glucose during exercise. Surprisingly, no differences were found in total CHO and fat oxidation between trained and untrained volunteers. However, although blood glucose oxidation rates were not different, exogenous CHO oxidation rates were higher in trained individuals. Differences in the results of van Loon et al. [55] and Burelle et al. [24] mayalsobecausedbydifferences in the experimental protocol (amount of CHO ingested, exercise intensity and timing of feedings). For instance, Burelle et al. [24] gave their first feeding (25g glucose) 30 minutes before exercise, which means that glycogen stores may have been pre-labeled, particularly in the trained volunteers who are more insulin sensitive and will have an increased muscle glucose uptake after an oral glucose load. This would result in an overestimation of exogenous CHO oxidation rates during exercise in the trained volunteers. If trained individuals stored 20% more of the initial glucose gift (5g) than the untrained individuals, this could explain the entire observed difference in exogenous CHO oxidation. This seems a reasonable assumption since it has been shown that post-exercise, glycogen resynthesis can be twice as fast after endurance training. [71] Three studies have investigated the effects of exogenous CHO oxidation during exercise at the same relative exercise intensity. [24,55,69] Two studies showed no effect of training on the oxidation of ingested CHO, whereas Burelle et al. [24] reported higher oxidation rates in trained compared with untrained individuals at 68% V. O 2max. The difference between these studies may be related to the fact that the latter study showed an increase in total CHO oxidation in trained individuals, whereas no changes in CHO oxidation were found in the studies by van Loon et al. [55] and Jeukendrup et al. [69] Burelle et al. [24] also reported increased muscle glycogen use, which is in contrast with most of the literature showing either no change or a decreased intramuscular glycogen breakdown after training at the same relative intensity. [61,62] In section 4 of this review we will discuss how maximal exogenous CHO oxidation rates are regulated. This concept, which is based on the premise that the liver and intestine play a crucial role in glucose homeostasis, describes that a maximal glucose output by the liver controls maximal exogenous CHO oxidation rates. This concept would predict that exogenous CHO oxidation rates are similar in trained and untrained individuals at the same absolute and relative workload. Higher exogenous CHO oxidation rates in trained individuals would suggest a superior absorption or more exogenous glucose would escape from the liver. There are currently no data available to support these potential differences between trained and untrained individuals. 4. Limitations of Exogenous CHO Oxidation As depicted in figure 3, exogenous CHO oxidation seems to be limited to rates of 1.0 to 1.1 g/min.

14 420 Jeukendrup & Jentjens CHO ingestion rate >2.0 g/min g/min Gastrointestinal tract 1.0 g/min? g/min Glucogen Glucose g/min 1.0 g/min 1.0 g/min 1.0 g/min Liver Blood Muscle CO 2 Fig. 5. Regulation of hepatic glucose production and the control of glucose appearance into the systemic circulation with carbohydrate (CHO) ingestion. CHO can be ingested at fairly high rates up to about 3 g/min before causing gastrointestinal symptoms. This CHO will then be digested and absorbed at a rate of 1.2to1.7g/min,whichhasbeensuggestedtobethemaximal absorptive capacity of the intestine. CHO will then enter the liver through the portal vein. A maximum of 1 g/min will escape from the liver and enter the bloodstream. The CHO entering the bloodstream may be derived from ingested CHO (in extreme conditions1g/min),canbederivedfromtheliver(glycogenolysis and gluconeogenesis) at a rate of 0 to 1 g/min, or can be derived from a combination of both. Whether glucose from ingested CHO can be directed towards liver glycogen during exercise has not been established. Glucose will be taken up by the muscle and can be oxidised at virtually similar rates. This graph was composed with results from various studies. [5,8,72] This finding seems supported by the vast majority of studies using either radioactive [3,36] or stable [5,37,45,53,69] isotopes to quantify exogenous CHO oxidation during exercise. One of the limiting factors could be gastric emptying. However, Rehrer et al. [31] showed that gastric emptying is unlikely to affect exogenous CHO oxidation rates. In their study, participants ingested 220g glucose during 80 minutes of exercise at 70% V. O 2max.After80minutes, 100g of glucose was present in the stomach and thus 120g was delivered to the duodenum. However, at 80 minutes only 38g of the ingested CHO was oxidised. These results were later confirmed by others using slightly different exercise protocols and feeding schedules. [27,38] Because in these studies only 32 to 48% of the CHO delivered to the intestine was oxidised, it was concluded that gastric emptying was not limiting exogenous CHO oxidation. Another potential rate-limiting factor is intestinal absorption of CHO. Studies using a triple lumen technique have measured duodenojejunal glucose absorption and estimated whole body intestinal absorption rates of a 6% glucose-electrolyte solution. [72] It was estimated that the maximal absorption rate of the intestine ranged from 1.3 to 1.7 g/min. Recent studies using stable isotope methodology have tried to quantify the appearance of glucose from the gut into the systemic circulation (Ra gut). When a low dose of CHO was ingested during exercise, the rate of appearance of glucose from the gut equaled the rate of CHO ingestion during the second hour (both 0.43 g/min). [5] This implies that at low ingestion rates absorption is not limiting and there is no net storage of glucose in the liver. Instead, all ingested glucose appears in the blood stream. It was also found that the glucose appearing in the bloodstream was taken up at similar rates to its Ra and 90 to 95% of this glucose was oxidised during exercise. When a larger dose of CHO was ingested (3 g/min), Ra gut was one-third the rate of CHO ingestion (0.96 to 1.04 g/min). Thus, only part of the ingested CHO entered the systemic circulation. However, the glucose appearing in the blood was taken up and 90 to 95% was oxidised. It was therefore concluded that entrance into the systemic circulation is a limiting factor for exogenous glucose oxidation, rather than intramuscular factors. This is further supported by glucose infusion studies. Hawley et al. [73] bypassed both intestinal absorption and hepatic glucose uptake by infusing glucose in volunteers exercising at 70% V. O 2max. When large amounts of glucose were infused and volunteers were hyperglycemic (10 mmol/l), it was possible to raise blood glucose oxidation rates above 1 g/min. These studies provide evidence that exogenous CHO oxidation is limited by the rate of digestion, absorption and subsequent transport of glucose into the systemic circulation rather than the rate of up-