Journal Logo

Basic Sciences: Original Investigations

Creatine Supplementation Reduces Muscle Inosine Monophosphate during Endurance Exercise in Humans

McConell, Glenn K.1,2; Shinewell, Joanna1; Stephens, Terry J.1; Stathis, Christos G.3; Canny, Benedict J.1; Snow, Rodney J.4

Author Information
Medicine & Science in Sports & Exercise: December 2005 - Volume 37 - Issue 12 - p 2054-2061
doi: 10.1249/01.mss.0000179096.03129.a4
  • Free


Over the approximate past 10 yr, much interest has been expressed in the exercise performance effects of creatine (Cr) supplementation. The balance of research in this area indicates that supplementing the diet with Cr, to the extent that an increase occurs in skeletal muscle total Cr, improves exercise performance during sprint exercise, particularly repeated sprints (5). In addition, Cr supplementation appears to increase strength and lean body mass in response to strength training programs (29). These improvements are thought to be caused largely by the increases in muscle total Cr, and particularly Cr phosphate, which makes a major contribution to adenosine triphosphate (ATP) resynthesis during such exercise (7,8). On the other hand, Cr supplementation does not appear to influence endurance exercise performance (1,2,28). Indeed, recently it was shown that 1-h time trial exercise performance is not influenced by Cr supplementation in trained cyclists (2). Interestingly, in that study the increases in plasma ammonia and plasma hypoxanthine during exercise were attenuated in the Cr supplementation trial, suggesting that a reduction in the extent of muscle energy imbalance may have occurred in the Cr trial (2). Muscle metabolite analysis, however, was not conducted in that study to directly test this possibility.

Both plasma ammonia (10) and plasma hypoxanthine (23) accumulation during exercise may reflect an energy imbalance in contracting skeletal muscle because they are produced in muscle when the rate of muscle ATP resynthesis fails to meet the rate of muscle ATP demand, resulting in an increase in muscle free AMP (6,10). Such an increase in skeletal muscle free AMP then increases the production of inosine monophosphate (IMP) and ammonia by AMP deaminase (10). Some of the ammonia can efflux from the muscle and enter the plasma (10). A small proportion of the accumulated IMP can undergo further degradation to produce inosine, which can be subsequently converted to hypoxanthine (6). A proportion of the hypoxanthine produced can also leave the contracting muscle and enter the circulation (23). Although the accumulation of ammonia and hypoxanthine in the plasma may reflect a muscle energy imbalance, issues cloud this interpretation. For example, during submaximal exercise, ammonia can also be produced in contracting muscle from the catabolism of amino acids and, therefore, the accumulation of plasma ammonia in this circumstance may not accurately reflect an energy imbalance (6). Therefore, measurements of key muscle energy metabolites such as the adenine nucleotides (ATP, ADP, AMP), IMP, Cr phosphate, ammonia, and lactate are more likely to provide an accurate indication of muscle energy status than plasma measures of ammonia and hypoxanthine.

Raising muscle total Cr content may alter muscle metabolism during aerobic exercise because of the central role of Cr in the muscle Cr phosphate shuttle (3,30). In the mitochondria, it appears that energy from ATP is transferred to Cr phosphate by mitochondrial Cr kinase. Energy is then transferred from Cr phosphate back to ATP in the cytosol by cytosolic Cr kinase such that ATP is made available at sites of high ATP use such as Na+/K+ ATPase, Ca2+ ATPase, and myosin ATPase (3,19). Increases in Cr phosphate in the cytosol may allow maintenance of a high ATP:ADP ratio at the Ca2+ ATPase (19) and other sites of ATP use (e.g., the actin-myosin cross-bridge interface) and, therefore, result in less of an increase in free AMP and, therefore, IMP during intense endurance exercise. Based on the findings of Bellinger et al. (2), the role of the Cr phosphate shuttle in skeletal muscle and the potential effects of increased Cr phosphate in the cytosol, we hypothesized that Cr supplementation would raise muscle total Cr levels and perhaps Cr phosphate and, therefore, reduce the extent of muscle energy imbalance during intense prolonged exercise lasting approximately 1 h. Based on previous studies examining endurance performance after Cr supplementation, we also hypothesized that performance would not be influenced by Cr supplementation. We chose to examine muscle responses after 45 min at approximately 80% V̇O2peak followed immediately by an approximately 15-min performance ride. This protocol allowed comparison of muscle and plasma metabolites between the control and the Cr supplementation experiment after a period of controlled matched exercise (first 45 min) and then after a performance ride where subjects were able to modify the power output to finish the required amount of work as quickly as they could. Although not the main focus of our study, we included an endurance exercise performance ride so that we could gain further insight into the factors influencing intense endurance exercise capacity. Therefore, if Cr supplementation attenuated the level of muscle energy imbalance during the performance ride, as hypothesized, but did not improve exercise performance, this would suggest that factors other than a moderate disturbance to muscle energy state influence exercise performance or capacity during prolonged intense endurance exercise.



Seven competitive endurance-trained cyclists or triathletes volunteered for this study, which was approved by the Monash University Standing Committee on Ethics in Research involving Humans. Before commencing the study, the subjects completed a medical questionnaire and provided informed, written consent. The age, height, and weight of the subjects was 21 ± 1 yr, 181 ± 3 cm, and 74.7 ± 2.1 kg (mean ± SEM), respectively. Competitive endurance trained subjects were chosen for this investigation due to their ability to perform reproducible efforts during exercise performance measurements.

Experimental procedures.

Approximately 2 wk before the first experimental trial, peak pulmonary oxygen consumption (V̇O2peak) was determined during continuous incremental cycling to volitional fatigue on an electromagnetically braked cycle ergometer (Excalibur Sport, Lode, Groningen, The Netherlands). The V̇O2peak averaged 4.83 ± 0.11 L·min−1 (64.8 ± 1.4 mL·kg−1·min−1). At least 3 d later, the subjects attended the laboratory for the first of two familiarization rides. For the 24 h before each familiarization ride, and the experimental trials, subjects were instructed to abstain from exercise, alcohol, and caffeine and to drink sufficient fluids such that their urine was “clear,” ensuring that they were euhydrated. In addition, the subjects were supplied with a standard breakfast (cereal, orange juice, and fruit; 1694 kJ), which they consumed 2 h before each familiarization ride and each experimental trial. The first familiarization ride involved a 5-min cycling warm-up (150 W) followed by 45 min of cycling exercise at a power output calculated from the subjects' V̇O2peak test to elicit approximately 80% of each participant's V̇O2peak (274 ± 7 W, 78 ± 1% V̇O2peak). Following this initial 45 min of exercise, the ergometer was placed into the “linear” mode function and each subject completed the equivalent amount of work as 15 min of cycling at 80% of each subject's V̇O2peak as quickly as possible. Because 1 W is equivalent to 1 J·s−1, we were able to calculate the amount of work (average 248 ± 7 kJ, range 225–272 kJ) that each subject, depending on their 80% V̇O2peak power output (274 ± 7 W), would complete in 15 min if they continued at the 80% V̇O2peak power output. Before the familiarization ride began, subjects were told their own personal performance ride work requirement that they needed to complete as quickly as possible following the first 45 min of exercise at 80% V̇O2peak. They were instructed to complete that amount of work as quickly as possible and, on average, they took 13.38 ± 0.45 min. Therefore, because the subjects were able to complete the work in substantially less than 15 min, this indicated that, on average, each subject was able to sustain a power output that was greater than 80% V̇O2peak (311 ± 11 W, calculated to be approximately 86% V̇O2peak). In the “linear” mode, an increase in pedaling rate elicits a greater power output, because W = L·(RPM)2, where L and RPM denote “linear factor” and revolutions per minute, respectively. To standardize motivational factors, no encouragement was given to the subjects during the performance rides of the familiarization rides or experimental trials, and the subjects were only permitted to see the accumulating kilojoules and not the clock or the instantaneous power output or RPM. As stated, this protocol allowed examination of muscle metabolism during approximately 1 h of intense endurance exercise involving a period of matched power output in the two experimental trials (first 45 min) and after a fixed amount of work was completed as quickly as possible (performance ride). This allowed greater insight to be gained in muscle metabolism and exercise capacity than could be obtained using a 1-h time trial type protocol because the power output is continually changing in that type of protocol.

At least 3 d later, the subjects completed a second familiarization ride, which was almost identical to the first familiarization ride (45 min: 279 ± 6 W, 80 ± 2% V̇O2peak; performance ride: 251 ± 6 kJ, range 234–272 kJ; 13.36 ± 0.32 min). The parameters of the second familiarization ride slightly differed from the first because minor modifications were made based on the first familiarization ride to ensure that the workloads reflected 80% V̇O2peak. The subjects were familiarized to the trials on two occasions in an attempt to reduce any learning effects during the experimental trials because we decided to administer the experimental trials in an ordered fashion. The washout time for muscle Cr (time for muscle total Cr to return to initial levels) following 5 d of Cr supplementation is approximately 4 wk (8). It was decided that 4 wk between trials was too long for comparative purposes when using trained subjects because it is likely that their training status or personal circumstances would change over that time period. Therefore, we decided that the control trial would be administered to all subjects first, followed the next week by the Cr trial. The subjects, however, were blinded to the order of the trials. No subject indicated after the study that they were aware of the order of the trials.

Over the 5 d preceding the first trial, the subjects ingested 140 mg·kg−1 body mass of dextrose four times per day (approximately 42 g·d−1 dextrose) (CON) then attended the laboratory 2 h after a standard breakfast and a final dextrose dose. Preceding the second trial, subjects ingested 70 mg·kg−1 body mass of dextrose and 70 mg·kg−1 body mass Cr monohydrate (CREAT) four times per day for approximately 5 d (21 g Cr monohydrate and approximately 21 g·d−1 dextrose), then attended the laboratory 2 h after a standard breakfast during which the final dose of Cr was ingested. The subjects were asked to thoroughly mix and dissolve the contents of each powder package in approximately 250 mL of warm water and to consume the contents of the packages evenly spaced throughout the day (e.g., breakfast, lunch, dinner, before bedtime). All subjects ingested all of the packages that they were given (as indicated on the form that each subject was given). On occasions, subjects forgot to take one of the packages but as instructed they then took two packages at the next ingestion time point.

Subjects reported to the laboratory in the morning at the same time for each trial. Trials were conducted 7 d apart. Subjects voided, and then nude body mass was measured to determine whether Cr supplementation increased body mass. A catheter was inserted into an antecubital vein and a blood sample was then obtained. The catheter for blood sampling was kept patent by flushing with 0.9% saline and after 30 min of exercise with approximately 0.5 mL of saline containing 10 U·mL−1 heparin. As was the case during the familiarization rides, the subjects completed a warm-up period involving 5 min of cycling at 150 W and then cycled for 45 min at 78 ± 1% V̇O2peak (279 ± 6 W), followed by completion of 251 ± 6 kJ as quickly as possible (performance ride). Using either the data from the two familiarization rides and the CON trial or the two familiarization rides and both the CON and CREAT trials yielded an identical coefficient of variation for performance time of 3.5 ± 0.7%. Jeukendrup et al. (9) also found a coefficient of variation of 3.5% for a very similar cycling protocol in well-trained athletes.

During the warm-up, and also at 15, 30, and 42 min of exercise, subjects consumed 3 mL·kg−1 body mass of water. Approximately 5 min before commencing exercise, after 45 min of exercise, and immediately after the exercise performance trial, muscle samples were rapidly obtained from the vastus lateralis muscle under local anesthesia using a needle biopsy technique, with suction (Bergstrom–Stille 5-mm biopsy needle). In preparation for the three muscle biopsy samples, three incisions were made in the skin and underlying fascia under local anesthesia approximately 3 cm apart, running proximal to distal in a line (one leg was used for the first trial with the other leg used for the second trial). The most distal site was used for the resting biopsy, the middle site for the 45-min sample and the proximal site for the final biopsy. Muscle samples were rapidly frozen in liquid nitrogen (<20 s from stopping exercise until muscle was frozen) for later analysis of muscle glycogen, lactate, ATP, ADP, AMP, IMP, ammonia, Cr phosphate (PCr) and Cr contents. A standard 1-min rest interval was assigned after the 45-min ride to conduct the muscle biopsy and prepare to commence the performance ride. Blood samples were collected just before the commencement of exercise (0 min), at 15, 30, and 45 min of exercise, and during the last 30 s of the performance ride. A portion of each blood sample was placed into a precooled tube containing lithium-heparin, which was immediately spun and the plasma transferred into a precooled cryotube, which was immediately frozen in liquid nitrogen for later analysis for plasma ammonia, Cr, and urate. Another portion of each blood sample was placed into a precooled tube containing fluoride-heparin, which was immediately spun with the plasma frozen at −20°C for later analysis of plasma glucose and lactate. The laboratory was maintained at 19–22°C and two fans, which were turned on at the same time in both trials, directed air over the anterior surface of the subject. Heart rate was recorded with a heart rate monitor (Accurex, Polar, Finland) and expired air was collected into Douglas bags for oxygen uptake and respiratory exchange ratio (RER) determination at 10 and 40 min of exercise. The subjects were asked to provide a rating of their perceived exertion (RPE) during exercise using a 14-point (6–19) Borg scale (4). The subjects were not given any feedback on their exercise responses until the completion of the second trial.

Analytical techniques.

Expired air samples were measured for oxygen and carbon dioxide content using Exerstress OX21 and CO21 electronic analyzers (Clinical Engineering Solutions, Sydney, Australia), which were calibrated using commercial gases of known composition. Expired air volume was measured using a dry gas meter (American Meter Company, Vacumed, Ventura, CA) calibrated against a Tissot spirometer. Plasma glucose and lactate were determined using an automated glucose oxidase and L-lactate oxidase method, respectively (YSI 2300 Stat, Yellow Springs, OH). Plasma ammonia was analyzed using a flow injection analytical technique as described by Svensson and Anfalt (27). Plasma hypoxanthine and urate were analyzed using high-performance liquid chromatography (HPLC) (25). Plasma Cr was analyzed using an enzymatic, fluorometric technique (12).

A portion (20 mg) of each muscle sample was freeze-dried then muscle lactate, PCr, and Cr were analyzed using enzymatic, fluorometric techniques and muscle ATP, ADP, AMP, and IMP were measured by HPLC as described previously (13,21). Although we freeze-dry each muscle sample and then crush it to a powder so that we can remove visible connective tissue, muscle biopsy samples contain variable amounts of nonmuscle fiber contamination (e.g., connective tissue, blood, fat) that can affect results. To minimize errors, we assume that the muscle sample from a set of muscle biopsies from a particular subject that contains the highest total Cr (Cr plus PCr) content is the sample that contains the least nonmuscle fiber contamination. The total Cr content is not affected by acute exercise. Because, in this study, the total Cr was expected to be higher in the CREAT trial than in the CON trial, we corrected the content of ATP, ADP, AMP, IMP, PCr, and Cr to the highest muscle total Cr content in each trial for each subject. We do not correct muscle glycogen (measured as glucose units after acid hydrolysis) and muscle lactate because of variable amounts of blood and, therefore, extracellular glucose and lactate in each muscle sample. Muscle ammonia (wet muscle) was analyzed using a flow injection analytical technique (27). Insufficient muscle was available for ATP, Cr, PCr, lactate, IMP, and glycogen analysis in two subjects at the 45-min time point in the creatine trial, and for muscle glycogen analysis at the end of the control trial in a different subject.

Statistical analysis.

Cardiorespiratory, blood, and muscle metabolite data from the two trials were compared using two-factor repeated measures ANOVA. If this analysis revealed a significant interaction, specific differences between mean values were located using the Fisher's least significance difference test. Performance time and pretrial body weight were compared using Student's paired t-tests. All data are presented as means ± SEM. The level of significance was set at P < 0.05.


Body weight, V̇O2, RER, heart rate, RPE, and performance.

Creatine supplementation had no significant effect (P > 0.05) on pretrial body weight (CREAT: 75.0 ± 1.8 kg; CON: 74.7 ± 1.7 kg). No differences in oxygen uptake and RER were observed between the two trials during exercise with mean values over the exercise period summarized in Table 1. Heart rate showed a significant (P < 0.05) trial by time interaction being on average 2–4 bpm higher in CREAT in the early stages of exercise but 2–5 bpm lower in CREAT during the performance ride (data not shown). At the end of the performance ride heart rate was 193 ± 4 in the CON trial and 191 ± 3 in CREAT (98 ± 1% in CON and 96 ± 1% in CREAT of the maximal heart rate obtained in the V̇O2peak test). RPE increased during exercise in both trials, with no significant differences between trials (data not shown). At the end of the performance ride, rating of perceived exertion was 18 ± 0 in both trials. Time to complete 251 ± 6 kJ in the performance ride was similar in the two trials (Table 1), and no significant difference was seen in the time to complete half the work (data not shown). Two subjects performed the two trials in essentially the same time (within 7 s), three a little faster in the CON trial, and two a little faster in the CREAT trial. The subjects paced the performance ride relatively evenly with no significant difference in the time taken to complete the first half of the 251 ± 6 kJ compared with the second half in either trial. The average power output during the performance rides was not significantly different between trials being 310 ± 8 W in CREAT and 313 ± 5 W in CON (both approximately 86% V̇O2peak), which were both significantly higher than the first 45 min of exercise (279 ± 6 W, 78 ± 1% V̇O2peak).

Mean physiologic responses during 45 min of exercise at 78 ± 1% V̇O2 peak and also time to complete 251 ± 6 kJ during the performance ride following 5 d of placebo (CON) or creatine monohydrate (CREAT) ingestion.

Blood measurements.

Plasma Cr concentration was significantly (P < 0.05) higher at rest in the CREAT trial compared with CON (Fig. 1A). During exercise, plasma Cr concentration increased from rest to 45 min and again from 45 min to the end of exercise in the CREAT trial, whereas it remained similar to rest during exercise in CON (Fig. 1A). A trial and time effect (with no interaction) for plasma urate indicated that plasma urate was lower in the CREAT at rest and during exercise with increases in urate during exercise in both trials (Fig. 1B, end higher than rest and 45 min in both trials). Plasma ammonia, hypoxanthine, and lactate concentrations were similar in the two trials at rest and increased (P < 0.05) to a similar extent during exercise in the two trials (Table 2). Plasma ammonia, hypoxanthine, and lactate concentrations increased markedly (P < 0.05) during the performance ride (Table 2). Plasma glucose concentrations were similar at rest and during exercise in the two trials (data not shown).

FIGURE 1— Plasma creatine (Cr) (A) and urate (B) at rest, following 45 min of exercise at 78 ± 1% V̇O2peak and after completion of the performance ride (251 ± 6 kJ) following 5 d of placebo (CON) or Cr monohydrate (CREAT) ingestion. Values are mean ± SEM;
FIGURE 1— Plasma creatine (Cr) (A) and urate (B) at rest, following 45 min of exercise at 78 ± 1% V̇O2peak and after completion of the performance ride (251 ± 6 kJ) following 5 d of placebo (CON) or Cr monohydrate (CREAT) ingestion. Values are mean ± SEM;:
N = 7. * Different from CON (P < 0.05). a Significantly different from rest (P < 0.05). b Significantly different from 45 min (P < 0.05).
Plasma ammonia, hypoxanthine and lactate concentrations at rest, during 45 min of exercise at 78 ± 1% V̇O2peak and after completion of the performance ride (251 ± 6 kJ) following 5 d of placebo (CON) or Cr monohydrate (CREAT) ingestion.

Muscle measurements.

Creatine supplementation resulted in a significantly (P < 0.05) higher muscle total Cr and PCr and significantly lower muscle ammonia (as indicated by significant treatment effects with no significant treatment by time interaction) (Table 3, Fig. 2). All of the subjects, except one, had higher TCr in CREAT than CON. Muscle Cr tended to be higher (P = 0.08) in CREAT than CON. Muscle PCr decreased and muscle Cr increased from rest to 45 min and then further to the end of exercise to a similar extent in the two trials (Table 3). The delta change in muscle PCr from rest to 45 min (22.0 ± 6.7 in CON and 33.0 ± 9.7 in CREAT) and from rest to the end of exercise (45.2 ± 7.1 in CON and 57.4 ± 5.9 in CREAT) were not statistically different between trials. The delta change in muscle Cr was almost identical to that of PCr. Muscle glycogen and lactate were similar at rest in the two trials and decreased and increased, respectively, to a similar extent in the two trials (Table 3). Muscle ATP and AMP were similar at rest and during exercise in the two trials with no significant change from rest observed in either trial (Table 3). Muscle ADP was not different between trials at rest or during exercise with the values at 45 min and the end of exercise being significantly higher than at rest in both trials (Table 3). Muscle IMP was similar at rest in the two trials and did not increase from rest to 45 min in either trial. At the end of the performance ride, muscle IMP increased in both trials with the increase in CREAT significantly less than in CON (Fig. 2A). The level of muscle IMP at the end of the performance ride was significantly higher in CON than CREAT with this being the case in all subjects. No relationship was seen between the level of muscle IMP at the end of exercise in the two trials and the trial exercise performance. A significant (P < 0.05) treatment effect and time effect were found for muscle ammonia, with a tendency (P < 0.09) for a treatment by time interaction (Fig. 2B). Muscle ammonia was significantly higher in both trials at the end of exercise compared with both at rest and at 45 min of exercise, with no significant difference in either trial between rest and 45 min of exercise (Fig. 2B).

Muscle metabolite contents (mmol·kg−1 dry muscle) at rest, following 45 min of exercise at 78 ± 1% V̇O2peak, and after completion of the performance ride (251 ± 6 kJ) following 5 d of placebo (CON) or creatine monohydrate (CREAT) ingestion.
FIGURE 2— Muscle inosine monophosphate (IMP) (A) and muscle ammonia (B) at rest, following 45 min of exercise at 78 ± 1% V̇O2peak, and after completion of the performance ride (251 ± 6 kJ) following 5 d of placebo (CON) or creatine monohydrate (CREAT) ingestion. Values are mean ± SEM;
FIGURE 2— Muscle inosine monophosphate (IMP) (A) and muscle ammonia (B) at rest, following 45 min of exercise at 78 ± 1% V̇O2peak, and after completion of the performance ride (251 ± 6 kJ) following 5 d of placebo (CON) or creatine monohydrate (CREAT) ingestion. Values are mean ± SEM;:
N = 7. * Different from CON (P < 0.05). # Significant treatment effect across the whole trial (P < 0.05). a Significantly different from rest (P < 0.05). b Significantly different from 45 min (P < 0.05).


The major finding of this study was that Cr supplementation, which had no effect on muscle IMP content during 45 min of exercise at 78% V̇O2peak, attenuated the increase in muscle IMP levels at the end of the subsequent performance ride (13.5 min at 86% V̇O2peak). We and others have interpreted an attenuation of the increase in skeletal muscle IMP during exercise as an indication that the muscle energy imbalance during exercise was lessened (13,16,20). Because intense endurance exercise performance was not influenced by Cr supplementation, despite evidence of a lessened muscle energy imbalance during exercise, these results suggest that factors other than a moderate disturbance in muscle energy balance may determine exercise capacity during such exercise.

Our interpretation depends on the validity of muscle IMP accumulation as a marker of impaired energy metabolism within muscle. When the ATP demand of the contracting muscle exceeds its rate of synthesis, levels of free ADP and free AMP increase, resulting in increased activities of myokinase and AMP deaminase, and ultimately an increased production of IMP (6). Because it is technically difficult to measure the free concentrations of ADP and AMP, muscle IMP accumulation has been used as a marker of metabolic stress (16,20). It is difficult to detect small decreases in muscle ATP content, given the relatively large store of this molecule, in contrast a small increase in muscle IMP accumulation above resting levels is more readily detected (24). That said, any IMP removal, either via reamination or degradation, will influence the muscle IMP levels and perhaps invalidate its use as a marker of an imbalance between the rates of ATP utilization and resynthesis during exercise. It is also important to keep in mind that measurements in mixed fiber muscle biopsy material may not reflect metabolic and functional changes at the muscle fiber level.

In this regard, it is important to note that in addition to a higher level of muscle IMP at the end of exercise in CON, muscle ammonia was also higher in the CON trial than in the CREAT trial (Fig. 2). This provides further evidence that AMP deamination to IMP and ammonia took place and that less occurred in CREAT than in CON. Ammonia is also produced by amino acid catabolism in muscle and it is likely that, at 45 min of exercise at 78 ± 1% V̇O2peak, a large proportion of the increase in muscle ammonia in both trials was caused by increases in muscle amino acid catabolism (6). This is suggested also by our data because no significant increase in muscle IMP occurred after 45 min of exercise in both trials (Fig. 2). During the performance rides, which were conducted at significantly higher exercise intensities (86% V̇O2peak, 310 W) than the first 45 min of exercise (78% V̇O2peak, 279 W), it is likely that the increase in muscle ammonia predominantly resulted from AMP deamination because a substantial increase in muscle IMP content was observed in both trials at this time (Fig. 2).

The reason for the apparent attenuation of the muscle energy imbalance during the performance ride in CREAT compared with CON is unclear. Creatine supplementation may have caused alterations in function in the mitochondria and/or the cytosol. One possibility is that the increase in total Cr content in the CREAT trial was associated with conditions in the mitochondria that improved oxidative processes. It is clear that ADP is an important stimulator of mitochondrial respiration but evidence also suggests that Cr stimulates respiration (11). In muscle, Cr kinase is located both in the cytosol and the mitochondrial intermembrane space. This has given rise to the Cr phosphate shuttle hypothesis (3,19) and the suggestion that Cr supplementation may improve both anaerobic energy production caused by the increased PCr levels and also aerobic metabolism by raising total Cr and, therefore, facilitating the Cr phosphate shuttle (2,18,30). Rico-Sanz (18) found some indications that skeletal muscle oxidative phosphorylation was enhanced during a single 200-s isometric plantar flexion contraction at 32% of maximal voluntary contraction after Cr supplementation in humans. In addition, ADP-stimulated mitochondrial respiration increases in permeabilized (skinned) human skeletal muscle fiber bundles when Cr is added and decreases when PCr is added (30).

Unlike Bellinger et al. (2), we found no effect of Cr supplementation on plasma ammonia or hypoxanthine during high-intensity submaximal exercise (Table 2). The data of Bellinger et al. (2), however, are difficult to interpret because, although plasma ammonia and hypoxanthine were higher during exercise before supplementation compared with after supplementation in the Cr supplemented group, the post creatine supplementation data were identical to the data obtained in a control group that did not receive Cr supplementation in either of their two trials. Similar to the present study, Balsom et al. (1) found no effect of Cr supplementation on plasma hypoxanthine concentration during prolonged endurance exercise.

As would be expected, plasma Cr concentration was greatly elevated before exercise in the CREAT trial, given the subjects ingested their last Cr dose approximately 2 h before attending the laboratory (Fig. 1). It was somewhat surprising, however, that plasma Cr concentration increased greatly (70%) during the exercise bout in the Cr supplementation trial, an increase that was obviously much greater than any possible shift in plasma volume (10% during a similar protocol, (15). No increase in plasma Cr concentration occurs after a 20-s sprint, which was preceded by 5 d of Cr supplementation (21). Plasma Cr concentration peaks approximately 1 h after the ingestion of a single 5-g dose of Cr and Cr concentration then decreases greatly from this value over the next hour (7). It is unlikely, therefore, that the increase in plasma Cr during exercise in CREAT was caused by high rates of absorption occurring more than 2 h after the ingestion. It is possible that plasma Cr increased during exercise, in part, because of a reduction in the rate of Cr excretion by the kidney, because the glomerular filtration rate has been shown to decrease during intense exercise (17).

Plasma urate concentration was lower in the Cr experiment, both before and during exercise, with a small but significant increase observed during exercise in both trials (Fig. 1). It is not known whether Cr supplementation lowers plasma urate concentration by reducing the production of urate or increasing the excretion of urate. Plasma urate is produced and released into the circulation from the normal breakdown of purines (adenine, guanine, hypoxanthine, xanthine), but also by direct synthesis from 5-phosphoribosyl pyrophosphate (5-PRPP and glutamine) within cells. In humans, two thirds of plasma urate excretion occurs through the kidney, whereas one third enters the gut where it is degraded (22).

Over the past few years, we have made attempts to determine the factors contributing to exercise capacity during exercise lasting approximately 60 min in endurance trained cyclists or triathletes. We have found that fluid ingestion, carbohydrate ingestion, alteration in blood acid-base balance, and now creatine supplementation have no significant effect on exercise performance (Table 1), (15,26) or exercise time to exhaustion (14). In our study examining the effect of carbohydrate ingestion versus placebo ingestion on exercise time to exhaustion at 83% V̇O2peak, we found no change in muscle PCr, IMP, and lactate from 32 min of exercise until exhaustion (70 min) in both trials (14). This suggests that exhaustion occurred in that study because of factors other than muscle metabolic energy balance. The present study supports that finding because Cr supplementation attenuated the rise in muscle energy imbalance during the performance ride but did not affect the ability to increase power output during approximately 15 min of exercise following 45 min of exercise at approximately 80% V̇O2peak. We anticipated that Cr supplementation would not influence endurance exercise performance because others have found no effect of Cr supplementation on prolonged (8–60 min) exercise performance (1,2,28). It should be noted, however, that a type 1 error may exist in our performance data because a small number of individuals was studied using a performance measure with a relatively large coefficient of variation of 3.5%.

In conclusion, creatine supplementation increased muscle total Cr and attenuated the increase in muscle IMP during an approximate 15 min performance ride (approximately 86% V̇O2peak) conducted following 45 min of endurance exercise in well-trained men. This suggests that raising muscle total Cr content before exercise can attenuate the extent of muscle energy imbalance during intense aerobic exercise. Because intense endurance exercise performance was not improved by Cr supplementation, despite lower muscle IMP levels, it is possible that exercise capacity during this type of exercise is determined by factors other than the muscle energy balance.


1. Balsom, P. D., S. D. Harridge, K. Soderlund, B. Sjodin, and B. Ekblom. Creatine supplementation per se does not enhanceendurance exercise performance. Acta Physiol. Scand. 149:521–523, 1993.
2. Bellinger, B. M., A. Bold, G. R. Wilson, T. D. Noakes, and K. H. Myburgh. Oral creatine supplementation decreases plasma markers of adenine nucleotide degradation during a 1-h cycle test. Acta Physiol. Scand. 170:217–224, 2000.
3. Bessman, S. P., and P. J. Geiger. Transport of energy in muscle: the phosphorylcreatine shuttle. Science 211:448–452, 1981.
4. Borg, G., G. Ljunggren, and R. Ceci. The increase of perceived exertion, aches and pain in the legs, heart rate and blood lactate during exercise on a bicycle ergometer. Eur. J. Appl. Physiol. Occup. Physiol. 54:343–349, 1985.
5. Casey, A., D. Constantin-Teodosiu, S. Howell, E. Hultman, and P. L. Greenhaff. Creatine ingestion favorably affects performance and muscle metabolism during maximal exercise in humans. Am. J. Physiol. 271:E31–E37, 1996.
6. Graham, T. E., and D. A. MacLean. Ammonia and amino acid metabolism in skeletal muscle: human, rodent and canine models. Med. Sci. Sports Exerc. 30:34–46, 1998.
7. Harris, R. C., K. Soderlund, and E. Hultman. Elevation of creatine in resting and exercised muscle of normal subjects by creatine supplementation. Clin. Sci. (Lond.) 83:367–374, 1992.
8. Hultman, E., K. Soderlund, J. A. Timmons, G. Cederblad, and P. L. Greenhaff. Muscle creatine loading in men. J. Appl. Physiol. 81:232–237, 1996.
9. Jeukendrup, A., W. H. Saris, F. Brouns, and A. D. Kester. A new validated endurance performance test. Med. Sci. Sports Exerc. 28:266–270, 1996.
10. Katz, A., S. Broberg, K. Sahlin, and J. Wahren. Muscle ammonia and amino acid metabolism during dynamic exercise in man. Clin. Physiol. 6:365–379, 1986.
11. Kuznetsov, A. V., T. Tiivel, P. Sikk, et al. Striking differences between the kinetics of regulation of respiration by ADP in slow-twitch and fast-twitch muscles in vivo. Eur. J. Biochem. 241:909–915, 1996.
12. Lowry, O. H., and J. V. Passonneau. A Flexible System of Enzymatic Analysis. New York: Academic Press, 1972,pp. 156–157.
13. McConell, G., R. J. Snow, J. Proietto, and M. Hargreaves. Muscle metabolism during prolonged exercise in humans: influence of carbohydrate availability. J. Appl. Physiol. 87:1083–1086, 1999.
14. McConell, G. K., B. J. Canny, M. C. Daddo, M. J. Nance, and R. J. Snow. Effect of carbohydrate ingestion on glucose kinetics and muscle metabolism during intense endurance exercise. J. Appl. Physiol. 89:1690–1698, 2000.
15. McConell, G. K., T. J. Stephens, and B. J. Canny. Fluid ingestion does not influence intense 1-h exercise performance in a mild environment. Med. Sci. Sports Exerc. 31:386–392, 1999.
16. Norman, B., A. Sollevi, L. Kaijser, and E. Jansson. ATP breakdown products in human skeletal muscle during prolonged exercise to exhaustion. Clin. Physiol. 7:503–510, 1987.
17. Poortmans, J. R. Exercise and renal function. Sports Med. 1:125–153, 1984.
18. Rico-Sanz, J. Creatine reduces human muscle PCr and pH decrements and P(i) accumulation during low-intensity exercise. J. Appl. Physiol. 88:1181–1191, 2000.
19. Rossi, A. M., H. M. Eppenberger, P. Volpe, R. Cotrufo, and T. Wallimann. Muscle-type MM creatine kinase is specifically bound to sarcoplasmic reticulum and can support Ca2+ uptake and regulate local ATP/ADP ratios. J. Biol. Chem. 265:5258–5266, 1990.
20. Sahlin, K., A. Katz, and S. Broberg. Tricarboxylic acid cycle intermediates in human muscle during prolonged exercise. Am. J. Physiol. 259:C834–841, 1990.
21. Snow, R. J., M. J. McKenna, S. E. Selig, J. Kemp, C. G. Stathis, and S. Zhao. Effect of creatine supplementation on sprint exercise performance and muscle metabolism. J. Appl. Physiol. 84:1667–1673, 1998.
22. Sorensen, L. B., and D. J. Levinson. Origin and extrarenal elimination of uric acid in man. Nephron 14:7–20, 1975.
23. Sorlie, D., K. Myhre, O. D. Saugstad, and K. E. Giercksky. Release of hypoxanthine and phosphate from exercising human legs with and without arterial insufficiency. Acta Med. Scand. 211:281–286, 1982.
24. Spencer, M. K., Z. Yan, and A. Katz. Carbohydrate supplementation attenuates IMP accumulation in human muscle during prolonged exercise. Am. J. Physiol. 261:C71–76, 1991.
25. Stathis, C. G., S. Zhao, M. F. Carey, and R. J. Snow. Purine loss after repeated sprint bouts in humans. J. Appl. Physiol. 87:2037–2042, 1999.
26. Stephens, T. J., M. J. McKenna, B. J. Canny, R. J. Snow, and G. K. McConell. Effect of sodium bicarbonate on muscle metabolism during intense endurance cycling. Med. Sci. Sports Exerc. 34:614–621, 2002.
27. Svensson, G., and T. Anfalt. Rapid determination of ammonia in whole blood and plasma using flow injection analysis. Clin. Chim. Acta 119:7–14, 1982.
28. Van Schuylenbergh, R., M. Van Leemputte, and P. Hespel. Effects of oral creatine-pyruvate supplementation in cycling performance. Int. J. Sports Med. 24:144–150, 2003.
29. Vandenberghe, K., M. Goris, P. Van Hecke, M. Van Leemputte, L. Vangerven, and P. Hespel. Long-term creatine intake is beneficial to muscle performance during resistance training. J. Appl. Physiol. 83:2055–2063, 1997.
30. Walsh, B., M. Tonkonogi, K. Soderlund, E. Hultman, V. Saks, and K. Sahlin. The role of phosphorylcreatine and creatine in the regulation of mitochondrial respiration in human skeletal muscle. J. Physiol. 537:971–978, 2001.


©2005The American College of Sports Medicine