Forty-nine healthy male subjects participated in this study. All subjects in this investigation participated in a familiarization session. During the familiarization session, subjects were informed as to the experimental procedures, completed a personal/medical history form, creatine supplementation history form and signed informed consent statements in adherence with the human subject's guidelines of the American College of Sports Medicine. The study was approved by the Ethical Review Committee of the University of Paderborn. Subject characteristics are presented in Table 1. No subject in this trial was a vegetarian with all subjects reportedly consuming meat in their daily diet. Exclusion criteria on admission were creatine supplementation within a period of 3 months prior to the experiments or the intake of any other nutritional supplement or medication at the time of the study. The subjects were instructed to avoid changes in their diet and training habits during the study.

A double-blind, placebo-controlled study was performed over a period of 5 weeks. The subjects performed 3 exercise tests at the same time of the day (day 1: incremental test, day 8: pre-test, day 35: post-test).

The subjects were assigned in random order to either the Cr-Pyr (n = 16), the Cr-Cit (n = 16) or the placebo (pla, n = 17) group. It was attempted to match groups for body mass/height ratio and the maximum weight achieved in the incremental test. Athletes with a high degree of handgrip utilization (martial arts, wrestling, etc) could possibly benefit to a greater degree due to training effect, as opposed to sports that neglect the upper extremity/hand grip (track, soccer, etc). To avoid potential differences between groups, athletes with a high degree of handgrip utilization have been equally distributed between groups. The 28-day supplementation period was started immediately after the pre-test and was continued until the day before the post-test. Subjects either received Cr-Pyr (Creapure™ Pyruvate, Degussa, Germany, containing 60% creatine and 40% pyruvate) or Cr-Cit (Creapure™ Citrate, Degussa, Germany, containing 65% creatine and 35% citrate) in form of lemon flavored effervescent tablets at a dose of 5 g per day, while the others received corresponding placebo supplements. Cr-Pyr and Cr-Cit contained <100 ppm creatinine, whilst dicyandiamide and dihyrdotriazine levels and polymeric pyruvates in CrPyr were not detectable by HPLC. The subjects were instructed to take the effervescent tablets (2 in the morning, 1 at noon and 2 in the evening) after the main meals accompanied by a carbohydrate rich drink. Intake of coffee and soft drinks containing caffeine was restricted to two cups a day.

Group data were expressed as mean ± SD and statistical significance was set at the p < 0.05 level. Subjects' characteristics before the investigation were compared using ANOVA. Pre and Post differences of anthropometric data and total means were analyzed by two-way ANOVA for repeated measures with post-hoc t-test considering the multi-comparison problem (Holm-Sidak). Force and relaxation velocity was analyzed by three-way ANOVA. In case of significant main effects for the treatment the effects within the groups were separately analyzed by a two-way ANOVA with repeated measurement in both factors.

Body weight increased in both creatine groups by a similar amount (p < 0.001, Table 1), whereas the weight in the placebo-group remained unchanged. Circumference of the forearm increased significantly in both creatine groups (p < 0.01, Table 1). Cr-Pyr and Cr-Cit were well tolerated. Creatinine levels increased significantly in both creatine groups but remained within the normal range. Data for all other measured blood parameters were in the normal range and no difference between groups could be determined.

The pre-test performance data was not statistically different among all groups. Supplementation with Cr-Pyr (p < 0.001) and Cr-Cit (p < 0.02) resulted in a significant increase in mean power (see Figure 3), while the placebo group showed no increase in power. A-vDiff in O₂ increased insignificantly (p > 0.05) in both creatine groups. The oxygen uptake to mean power ratio was slightly reduced with Cr-Cit and Cr-Pyr, however, the changes were not statistically significant (p > 0.05, Figure 3). Blood flow was enhanced with Cr-Pyr but decreased slightly with Cr-Cit, however, both were not statistically significant (p > 0.05). A-vDiff in O₂ and blood flow measured during the recovery periods significantly increased (p < 0.05) in the Cr-Pyr group concordantly with muscular oxygen consumption compared to the pre-tests. Venous blood lactate levels remained unchanged with either treatment. Mean muscular NH₃ release was significantly reduced in the Cr-Cit group (p < 0.05). The NH₃ release per unit of power decreased after Cr-Pyr and Cr-Cit treatments, conversely, only Cr-Cit group resulted in significant changes (p < 0.015, Figure 3).

The contraction velocity increased 24% with Cr-Pyr (p < 0.01) and 20% with Cr-Cit (p < 0.01, Table 2). This increase is accompanied by a significant increase in force (p < 0.001) and relaxation velocity (p < 0.01) in the Cr-Pyr group. The Cr-Cit group showed a significant increase in force (p < 0.01), however, the change in relaxation velocity was not significant. In addition, Cr-Cit significantly increased force only during the first two intervals, when compared to placebo. The effect of Cr-Cit on force decreased over time and the improvement was not significant at interval 6 and 9. The increase in contraction frequency is the major reason for the increase in mean power.

The results of this study suggest that 4 weeks of low dose Cr-Pyr and Cr-Cit intake significantly increased mean power in comparison to placebo. Cr-Pyr improved contraction velocity and reduces fatigability during intermittent exercise of high intensity. Cr-Pyr showed significant improvements in force during all intervals, whereas the effects of Cr-Cit decreased and improvements were not significant during the later intervals. The effect of Cr-Pyr resulted from an increased contraction and relaxation speed and is accompanied by enhanced oxygen consumption and blood flow.

Daily creatine intake ranged from 0.027 to 0.047 g/kg bw (mean 0.037) in the Cr-Pyr group and from 0.036 to 0.052 g/kg bw (mean 0.044) in the Cr-Cit group. Despite a 19% difference in mean creatine dosage per body weight there was a similar increase in body weight in both creatine groups without significant changes in body fat. Weight increased in 15 of 16 subjects (94%) with Cr-Pyr and in 14 of 16 subjects (88%) with Cr-Cit. The weight gain is similar to previously reported weight gains of short-term, high-dose 629 or long-term, low-dose creatine supplementation 12.

The efficacy of creatine during the first exercise interval is dependent on the type, the intensity and the duration of exercise. The exercise must be of high intensity and of a certain duration allowing PCr stores to drop significantly. Additionally, the efficacy of creatine supplementation seems to be dependent on the speed of movement. During isometric exercise, creatine supplementation was found to have no or small effects on isometric torque production 3031. An increase in power of about 20% was measured in fast movements using Wingate tests 32 and isokinetic tests resulted in moderate improvements of mean and peak power (6% and 8%, respectively) 33. Cr-Cit and Cr-Pyr supplementation increase performance during the first intervals (see Table 2) in contrast to previously published data 534. The increase of Cr-Pyr group was larger compared to the Cr-Cit group; however, the difference was not significant. The type of exercise used in this study consisted of a large amount of negative work. If only the active work is considered (contraction velocity, as the reduction of the displacement can be neglected in the first interval), the improvement is 18 and 15%, respectively, an increase comparable to studies using Wingate tests.

The increase in performance in both creatine groups at the beginning of the intermittent exercise is related to an increased contraction speed. Contraction speed during fatiguing exercise seems to be reduced when free ADP increases 35. Creatine has been reported to increase ATP re-synthesis and thus reduce the increase in free ADP 36. This is confirmed by smaller muscular NH₃ release during exercise which suggests a lower adenosine (AMP) to inosine (IMP) conversion rate (ATP dependent), and thus lower inorganic phosphate concentration. A reduced NH₃ release or concentration in blood was shown in previous creatine supplementation studies 3337. Creatine supplementation results in reduced hypoxanthine concentrations 37 which might be the explanation for the observed improvement in the first interval with Cr-Cit and Cr-Pyr. Relaxation speed increased in both Creatine groups after the first exercise interval comparable to previous observations 31. But the increase was only significant with Cr-Pyr.

The effect of creatine supplementation on performance in repeated intervals varies with the diversity of tests and the beneficial effect decrease at later intervals of the exercise. Short-term, high-dose creatine supplementation did show the greatest improvement in performance between intervals 4 and 7 during 10 intervals of 6 sec maximum cycling exercise 34. Subsequent intervals showed a decrease in improvement 34. Thirty sec exercises intervals resulted in a decrease of improvement in performance from the second interval 36, showed no improvement in the third interval 33, or no difference in the fourth and fifth interval 5. The beneficial effect of creatine intake on interval exercise was related to an increased rate of PCr re-synthesis during breaks 638 but the existing literature is inconclusive. PCr re-synthesis was found not to be enhanced 38 or even reduced 39 and the rate of synthesis defined as dPCr*s^-1 after dynamic 38 and isometric work 31 was not influenced by creatine supplementation. Cycling with maximal intensity for 30 sec resulted in decreased PCr levels of 49 mmol kg^-1 dry weight without and 57 mmol kg^-1 dry weight with creatine ingestion 36. However, the difference in re-synthesis after exercise was only 2 mmol after 4 min 36 indicating an enhanced accumulation of inorganic phosphate (Pi) during repeated intervals, especially if the breaks are short. Increased Pi might cause fatigue 40 which might be compensated by an increased PCr pool during early intervals. PCr decreases during later intervals 33 and is not able to compensate the effects of increased Pi. The decrease in performance relative to the initial performance might be greater after creatine supplementation 5313336. Cr-Cit intake showed such a performance pattern (see Table 2) indicating that fatigue occurred in parallel to PCr-breakdown 41. If PCr re-synthesis would have been increased after creatine ingestion, Pi concentration should not increase. As a result, ATP production must be enhanced resulting in an increased oxygen uptake during the breaks. Comparable to previous research did Cr-Cit supplementation not change oxygen uptake. Therefore the ratio VO₂ to mean power decreased slightly (see Figure 3), indicating that the additional energy was derived from anaerobic sources and that PCr re-synthesis was not enhanced.

The increase in performance with Cr-Pyr during the first intervals is slightly larger than with Cr-Cit. Moreover, Cr-Pyr reduces fatigability during all following intervals. The improvement results from an increase in contraction velocity during all intervals and an increase in relaxation velocity 31. This suggests ADP or Pi levels are reduced and ATP re-synthesis is increased during exercise, which should lead to an increased VO₂. In fact, the improvement in performance is accompanied by an increase in muscular VO₂ measured during the breaks. As blood flow decreases during the 15 sec of exercise due to the high forces occurring 42, muscular oxygen uptake during exercise cannot be very high. Therefore, measurements during the breaks are representative for the oxygen consumption during the exercise period. The ratio VO₂ to mean power drops similar to Cr-Cit indicating that most of the improvement is due to the increased VO₂, however a small amount of energy is still derived from anaerobic sources. The energy must be derived from phosphate stores since lactate levels did not change. The amount is comparable to that observed after the Cr-Cit supplementation. The additional amount of aerobic work after Cr-Pyr supplementation compensates the creatine effect on ammonia release as the ratio ammonia release to mean power is not significantly different between pre and posttests. Thus the improvement by Cr-Pyr seems to be caused by a different mechanism perhaps resulting in increased ADP concentration compared to Cr-Cit.

A potential creatine effect on the time course of VO₂ is related to Phosphocreatine breakdown (Bessman cycle or Phospho-Creatine shuttle) 43. Cr-Cit and Cr-Pyr supplementation should result in a similar increase in muscle creatine content; hence varieties in VO₂ cannot be explained by differences in the Bessman cycle. Additionally, there are some studies showing that VO₂ during exercise is unaffected 9, or the VO₂-kinetics is even slowed down after creatine supplementation 44 and oxygen consumption after exercise of high intensity or during breaks is unchanged 45.

The observed results suggest additional pyruvate benefits in the Cr-Pyr group. The role of pyruvate supplementation might be to decrease the relative inhibition on aerobic glycolysis due to elevated creatine phosphate and ATP levels. It may also be reflective of a more rapid regeneration of ATP and reduction in the inorganic phosphate increase, allowing for higher intracellular calcium ion concentration, permitting the higher frequency contraction and relaxation. To achieve an additional effect of pyruvate an increase in plasma pyruvate concentration should occur. A recent study showed that one-time administration of 5 g Cr-Pyr does not increase plasma pyruvate concentrations 21. However, previous research found a cumulative effect of pyruvate administration increasing the plasma pyruvate concentration by about 60% after 4 weeks 18. In the intestine and the liver, pyruvate can easily be converted to ketone compounds 46 or different amino acids 47. If parts of these substances or of the pyruvate absorbed were distributed to the muscle the pool of substances for anaplerotic reactions would increase. This enables the muscle to replenish the pool of tri-carbon-cycle intermediates rapidly at the beginning of exercise. This replenishment might favor the flux through the TCA cycle and thus increase the delivery of protons to the respiratory chain. The increased concentration of intermediates might also act as a pyruvate buffer, with the function to deliver carbon skeletons to the TCAC. Both mechanisms could be beneficial during the transition from rest to high intensity exercise to increase oxygen uptake 48. Given the transient presence of anaplerotic substrates, the effect might be short-lived and not subject to long-term accumulation over time. The results of more acute administration or different organic salts of pyruvate would be of interest.