INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CAFFEINE AND CREATINE are popular ergogenic substances in athletic populations. Both substances occur in the normal diet yet are ingested as supplements by athletes for the express purpose of enhancing exercise performance (for recent reviews on these issues, see Refs. 22 and 24).

It is well established that caffeine, even at dosages that are well below the reference limit set by the International Olympic Committee, can enhance endurance exercise performance (22). Investigations examining its effects in short high-intensity activities requiring strength and power have produced inconsistent results. Caffeine enhanced maximum power output in 6-s sprints (1) but not in a 30-s Wingate test (5, 13). In a study by Greer et al. (13), healthy volunteers performed four 30-s Wingate sprints with 4-min rest intervals in between. Caffeine did not impact on power output during the first 2 sprints, but it was detrimental to performance in the latter 2 exercise bouts. Findings with regard to the impact of caffeine on strength activities are even more controversial. However, this is at least partly due to lack of well-controlled studies addressing this issue. Interestingly, Kalmar and Cafarelli (16) recently reported caffeine to increase time to fatigue during voluntary knee extensions at 50% of maximal voluntary contraction (MVC) force by ~25%, whereas voluntary activation increased by 3.5%. Neither the force-electromyogram relationship nor motor unit firing rates were altered by caffeine, which indicates that the ergogenic effects of caffeine were, at least in part, mediated by local mechanisms. Accordingly, two other studies found caffeine intake in humans to potentiate muscle force production during low-frequency nerve stimulation (18, 23).

More than 50 studies over the last 10 yr have addressed the effects of creatine supplementation on high-intensity exercise performance. Data consistency has been curtailed by results of poorly controlled studies. However, the body of evidence clearly indicates that short-term creatine supplementation (20-30 g/day, 4-7 days) can increase maximal force and power output during short maximal exercise bouts, in particular during intermittent exercise but not during endurance exercise (24). Furthermore, creatine supplementation in conjunction with heavy resistance training has the potential to stimulate muscle hypertrophy and maximal muscle force (27). This effect is important to any sport involving heavy resistance training.

The physiological mechanisms by which caffeine and creatine elicit their ergogenic effects are poorly understood. The failure of central neural as well as systemic metabolic mechanisms to explain the effects of caffeine in short maximal exercise has prompted interest in its local actions on muscle cells. Our laboratory has demonstrated in perfused rat muscles that physiological concentrations of caffeine stimulate glycogenolysis in glycolytic muscle fibers during electrical stimulation (29). However, glycogen availability is not limiting high-intensity exercise performance. In a recent study, Tarnopolsky and Cupido (23) demonstrated caffeine ingestion to potentiate muscle force production during low-frequency tetanic stimulation. A putative mechanism to explain this finding is that caffeine at physiological concentrations facilitates sarcoplasmic reticulum calcium release. The beneficial effect of caffeine intake on performance during electrically induced cycling in tetraplegic patients also indicates a direct action of the drug on skeletal muscle (20). In fact, our laboratory's earlier finding that caffeine intake can negate the ergogenic effect of oral creatine supplementation (26) also supports a local mechanism of action of caffeine. The exact cellular mechanisms linking increased intracellular creatine content in muscle due to creatine supplementation (14, 27, 28) with enhanced contractile performance (12, 26, 27) remain largely unexplained. However, it is obvious that increased creatine uptake by muscle fibers during creatine supplementation initiates intracellular responses that eventually result in ergogenic effects.

Our laboratory has recently reported that creatine supplementation shortened relaxation time (RT) during intermittent maximal isometric muscle contractions (25). Sarcoplasmic reticulum Ca²⁺ reuptake by virtue of Ca²⁺-ATPase pump activity is the rate-limiting step in relaxation of mammalian muscle cells (7, 11). Thus the effect of creatine intake on relaxation rate provides evidence to suggest that sarcoplasmic reticulum Ca²⁺ reuptake is facilitated in creatine-loaded muscle. If extracellular caffeine concentrations established by oral caffeine intake are indeed effective to enhance Ca²⁺ release from the sarcoplasmic reticulum (23), then caffeine administration conceivably might undo the effect of creatine supplementation to facilitate muscle relaxation. Moreover, because caffeine can counteract the ergogenic effect of creatine supplementation, such an observation would provide indirect evidence to support our hypothesis that shortening of muscle relaxation is important to the ergogenic impact of creatine supplementation during sprint exercise.

Therefore, in the present study we investigated the effects of creatine and caffeine supplementation, alone or in combination, on muscle RT. A recent study in isolated rat hearts has shown that the adaptations of contractile function may markedly differ depending on the duration of caffeine exposure (17). Therefore, we compared the effects of acute and short-term (3 days) caffeine administration. Furthermore, to eliminate neural effects, we studied muscle RT during electrically induced contractions. This approach is valid because the rate-limiting step of muscle relaxation, notably sarcoplasmic reticulum Ca²⁺ reuptake, is independent of whether contractions are either voluntary or induced by electrical stimulation (7, 11). We used isometric contractions because doing so yields highly reproducible measurements of muscle RT (25).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Subjects. Ten physical education students, 9 men and 1 woman, aged between 21 and 24 yr, gave their informed, written consent to take part in the study. All subjects were physically active, but none of them was highly trained. They were informed of the experimental procedures to be undertaken and were instructed to abstain from any medication and caffeine starting 1 wk before each experimental period (28). Furthermore they were asked to avoid changes in their usual level of physical activity and diet during the entire study period lasting 25 wk. The local ethics committee approved the study protocol.

Study protocol. A double-blind study was performed, whereby the subjects were assigned in random order to five experimental protocols, each lasting 8 days (day 0 to day 7) and separated by a washout period of 5 wk. There is good evidence to suggest that muscle creatine stores return to normal within 4-5 wk after cessation of creatine supplementation (27). One week before the start of the study, the subjects reported to the laboratory to get habituated to the test protocol. The protocol involved electrical stimulation of the left musculus quadriceps femoris and aimed to assess muscle activation time, isometric torque, and RT. The subjects returned to the laboratory for the pretest on day 0. One week after the pretest (day 7), at the same time of day, the subjects returned to the laboratory for the posttest, the procedures of which were identical to the pretest. The results of the measurements were disclosed neither to the subjects nor to the investigators until completion of the entire study. In protocol A (Cr) the subjects ingested 4 × 5 g of creatine monohydrate per day for 4 days (day 3 to day 6). It is well established that this mode of creatine supplementation is effective to raise muscle creatine content in young male volunteers (14, 24). The last creatine dose preceded the measurements on day 7 by at least 12 h. In protocol B (Caf), on days 4, 5, and 6 the subjects received a single dose of 5 mg caffeine · kg body wt¹ · day¹. The caffeine capsules were ingested at breakfast. The last dose of caffeine preceded the measurements on day 7 by at least 20 h. Protocol C (Cr+Caf) was a combination of protocols A and B. Thus, on days 4, 5, and 6, in addition to the creatine supplements, the subjects were administered a single dose of 5 mg caffeine · kg body wt · ¹ · day¹. Similar to protocol B, the caffeine capsules were ingested at breakfast together with a creatine supplement. Our laboratory had previously demonstrated that such combination of creatine and caffeine counteracts the beneficial impact of creatine supplementation alone on muscle power output during maximal intermittent muscle contractions (26). Protocol D was an acute caffeine condition (ACaf): the subjects received 5 mg caffeine/kg body wt 1 h before the exercise test on day 7 only. In protocol E, the subjects ingested placebo supplements from day 3 to day 7. The creatine powders were flavored by the addition of citrate (60 mg/g creatine) and maltodextrine (940 mg/g creatine). The placebo powders contained maltodextrine and citrate (40 mg/g maltodextrine). The subjects were instructed to dissolve the powder supplements in ~150 ml of hot water immediately before ingestion. Creatine and placebo powders were identical in taste and appearance. Placebo capsules contained maltodextrine. The administration schedule of the powder supplements (creatine or placebo) and capsules (caffeine or placebo) was identical for all experimental conditions.

Electrical stimulation protocol. The subjects arrived at the laboratory in the morning between 8:00 and 12:00 AM. First their body weight was measured. They were then seated on a chair, rotated 30° backward, with the left leg supported at the level of the popliteal space. An isokinetic dynamometer, instrumented with a torque transducer (Lebow 1605, 0.05% accuracy level) and connected with a rigid lever arm, was positioned lateral to the knee such that its axis was aligned with the knee joint axis. The leg was tightly strapped to the lever arm of the system proximal to the ankle joint. The lever arm was rotated to a knee angle of 90°. The positions of the knee support and of the axis of the isokinetic dynamometer and the length of the lever arm were noted for each individual, to be set identically for the pre- and posttests for all experimental conditions. Two electrodes (6 × 4 cm) were then fitted on m. quadriceps femoris. The proximal electrode was positioned at one-third of the line connecting the anterior superior spina iliaca and the center of the upper edge of the patella, with the knee in full extension (180°). The distal electrode was positioned 15 cm distal to the proximal electrode. After a standardized 5-min warmup, the subject performed three voluntary maximal isometric knee extensions (3 s) at a knee angle of 90°, interspersed with 10-s rest intervals, to determine MVC. Thereafter, the rectus femoris muscle was electrically stimulated (Multistim ESP, Fysiomed, Edegem, Belgium) with 2-s trains, composed of 0.5-ms unipolar square-wave pulses firing at 60 Hz. Over a maximum of 5 contractions (10-s rest pauses), the stimulation intensity was gradually increased to obtain the target force of 20% of the previously determined maximal voluntary isometric knee-extension torque, or less in case of excessive pain (n = 2, force = ~15% of MVC). After 5 min of rest, 30 electrically induced contractions (2 s) were initiated with a 2-s rest-interval between the contractions. For each contraction, both torque and the activation pattern of the electrical stimulator were digitized at 500 Hz and stored to disk for later analysis. Stimulation intensity was on average 95 mA (range 56-110 mA) and was identical for the five experimental conditions.

Data analysis. The torque from each contraction was characterized by three parameters: 1) maximal torque (T_max, in N · m); 2) contraction time (CT, in ms), which was defined as the time needed to increase muscle torque from 0.25 to 0.75 of T_max, and 3) RT (in ms), which was defined as the time needed to reduce torque from 0.75 to 0.25 of T_max. Our laboratory previously reported that the intraday reliability for CT and RT measurements with the present methodology is 0.94 and 0.96, respectively. Corresponding values for interday reliability were 0.92 and 0.95 (25). The subjects performed 30 contractions induced by electrical stimulation. Mean CT, T_max, and RT were calculated for the entire series of 30 contractions, as well as per six contractions (1-6, 7-12, 13-18, 19-24, 25-30) to consider the effects of fatigue. Treatment effects were evaluated by either two-way (means of 30 contractions) or two-way (means per 6 contractions) repeated-measures analyses of variance, which were covariate adjusted for the pretest values. Tukey's tests were used as a post hoc test to locate differences among means. The level of statistical significance was set at P  0.05. All statistical procedures were performed with use of Statistica software (Statsoft, Tulsa, OK).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Maximal knee-extension torque. In placebo, initial maximal knee-extension torque (T_max; mean of contractions 1-6) was 28 ± 3 N · m in the pretest and 30 ± 3 N · m in the posttest (Table 1). T_max decreased (P < 0.05) by ~40% from the start to the end of the contraction series. Thus final T_max (mean of contractions 25-30) was 17 ± 2 N · m during the pretest vs. 19 ± 3 N · m during the posttest. T_max was similar between all experimental conditions throughout the contraction series, during both the pretest and the posttest. Immediately before the start of the electrical stimulation protocol, MVC was measured. MVC in placebo was similar in the pretest (157 ± 10 N · m) and in the posttest (159 ± 10 N · m), and there were no differences between the experimental conditions at any time (data not shown). Thus initial T_max during electrical stimulation consistently amounted to 18-20% of MVC.

CT. Initial CT in placebo was 79 ± 4 ms during the pretest vs. 80 ± 4 ms during the posttest (Table 2). CT lengthened (P < 0.05) with increasing number of contractions. At the end of the contraction series, CT was increased to 91 ± 4 and 94 ± 3 ms in the pretest and posttest, respectively. There were no significant differences for CT between the treatments at any time.

RT. Initial RT was 58 ± 3 ms in the pretest and 60 ± 3 ms in the posttest (Table 3). RT gradually increased (P < 0.05) throughout the stimulation protocol. RT during the final six contractions of the series was on average 86 ± 5 and 89 ± 4 ms during the pretest and posttest, respectively. Furthermore, RTs at the pretest were similar for all experimental conditions. Compared with placebo, during the posttest mean RT was lower in Cr (P < 0.05) and tended to be higher in Caf (P = 0.09). Compared with Cr, short-term caffeine intake, alone (Caf) or in conjunction with creatine intake (Cr+Caf), resulted in longer RTs. This lengthening of RT during Caf and Cr+Caf was most prominent during the final stage of the contraction series (Table 3). RT was not significantly different at any time between Caf and Cr+Caf. During ACaf, mean RT was not significantly different from placebo. However, on the one hand RTs during ACaf were longer than during Cr (P < 0.05); on the other hand they were shorter than during Caf (P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study confirms that oral creatine supplementation shortens muscle RT in humans (25). During a series of 30 muscle contractions induced by intermittent electrical stimulation, RT after creatine supplementation was reduced by 5% and was significantly shorter than after placebo. RTs exhibited the normal increase with fatigue (8) in either group, yet they were consistently lower during creatine supplementation than during placebo (Table 3). Furthermore, consistent with our earlier observations, the shortening of RT occurred in the absence of any change of either peak force production (Table 1) or CT (Table 2). Interestingly, the intake of caffeine (5 mg/kg body wt) together with one of the daily creatine supplements for a period of 3 days counteracted the beneficial effect of creatine intake on RT, and this inhibitory effect was enhanced by fatigue. The fatigue-induced lengthening of RT was greater after caffeine intake than during placebo. Slowing of relaxation in fatigued muscle has been explained by inhibition of sarcoplasmic reticulum Ca²⁺-ATPase activity because of falling pH and ATP affinity (8, 11). Our findings thus suggest that caffeine reduces the functional capacity of sarcoplasmic reticulum Ca²⁺ ATPase.

It is intriguing that short-term caffeine supplementation (5 mg · kg¹ · day¹ for 3 days before the posttest) but not acute caffeine intake (5 mg/kg bolus 1 h before the posttest) impaired muscle relaxation. The last dose of caffeine during the Caf protocol preceded the experiments by at least 20 h. Hence plasma caffeine was largely washed out at the time relaxation rates were measured (3). Our present data thus clearly show that the effects of caffeine on muscle relaxation can be different depending on the administration regimen used. In keeping with such contention, in a recent study 8 wk of caffeine treatment (10 mg/kg ip, twice per day) in rats was found initially to decrease myocardial contractility yet thereafter to increase it (17). The mechanism underlying this time phenomenon cannot be explained from the present findings. Caffeine at physiological concentrations (<75 µM) is known to exert its actions primarily by inhibiting cellular adenosine receptors (9, 10), and A₂ subtype adenosine receptors are present in the plasma membrane and cytosol of muscle cells (19). In addition, adenosine has been found to be involved in relaxation of vascular smooth muscle cells by acting on intracellular A₂ adenosine receptors (4). Furthermore, studies in rats have shown that chronic dietary caffeine administration not only can markedly increase plasma adenosine concentration (6) but also can change adenosine receptor action in peripheral cells (30). However, the impact of short-term caffeine intake and withdrawal on adenosine receptor distribution and signaling in human skeletal musculature is at present unknown.

It has been previously demonstrated in humans that caffeine intake at a rate (4-7 mg/kg body wt) that corresponds to the present dosage (5 mg/kg body wt) can potentiate contraction force in fatigued muscles during low-frequency electrical stimulation (20 Hz) but not during high-frequency stimulation (40 Hz; Ref. 18, 23). This observation can probably be explained by the fact that physiological caffeine concentrations can partially counteract (21) the impairment of calcium release from the sarcoplasmic reticulum ryanodine receptor occurring in muscles fatigued by low-frequency electrical stimulation (2). Conversely, muscle fatigue during electrical stimulation at higher frequencies has been attributed to reduction in electrical activity of the muscle, which means proximal to the ryanodine receptor (2). Consistent with the above observations, during the high-frequency (60 Hz) electrical stimulation protocol used here caffeine did not impact on either the rate (CT) or magnitude (T_max) of muscle tension development either in the absence or presence of fatigue (Tables 1 and 2).

The exercise model used in the present study was not suitable to evaluate the ergogenic effects of creatine supplementation. Creatine supplementation has been found to be effective to improve performance during repetitive high-power output exercise bouts (24) but not during isometric muscle contractions (25, 26). Nevertheless, we used isometric contractions for the express purpose of obtaining highly reliable measurements of muscle RT (25), which in our hands is impossible during dynamic muscle contractions (Van Leemputte, unpublished observations). Still, our findings provide indirect evidence to suggest that facilitation of muscle relaxation may be important to the ergogenic action of creatine supplementation. First, we clearly show here that caffeine negates the beneficial effect of creatine loading on muscle RT. Second, our laboratory has previously reported that short-term caffeine administration, according to an identical intake regimen as used here, entirely counteracted the ergogenic effect of creatine loading in maximal intermittent muscle contractions (26). Furthermore, from a theoretical point of view muscle relaxation rate is conceivably important to power production during "sprint" exercise. First, during fast repetitive concentric muscle contractions, recovery time from the previous contraction is very critical to maximal force output during the next contraction (15). Second, it is reasonable to assume that shortening of RT might increase the number of actomyosin activation cycles per unit of time and thereby power output.

Finally, it is important to mention that muscle creatine content was not measured in the present study. However, the creatine and caffeine supplementation regimens used here were adopted from an earlier study by our laboratories in similar subjects (26). In the latter study, creatine alone and creatine plus caffeine administration caused a similar ~5% increase of muscle PCr content assessed by ³¹P-NMR spectroscopy. Thus the differential impact of creatine and caffeine on muscle relaxation is not related to differences of intracellular creatine content in muscle cells.

In conclusion, the present study shows that short-term caffeine intake, but not acute caffeine intake, inhibits muscle relaxation. This negative impact of caffeine on RT counteracts the beneficial effect of creatine supplementation on muscle RT.