INTRODUCTION

TOP ABSTRACT INTRODUCTION RESEARCH DESIGN AND METHODS RESULTS DISCUSSION REFERENCES

THE MAJOR PART OF THE WHOLE BODY insulin-induced glucose disposal is directed toward the skeletal musculature. One of the factors that can alter insulin sensitivity in human skeletal muscle is the degree of muscular activity, which is reflected by changes in the expression level of the glucose transporter (GLUT-4) as well as the glycogen storage capacity (23). An acute reduction in activity level, such as occurs in leg immobilization or bed rest, is known to rapidly decrease muscle insulin sensitivity and glucose transporter (GLUT-4) content (24, 27). The opposite is true in the case of an increased activity pattern, such as training (6, 14). If a significant portion of the body's musculature underwent an increase in habitual activity level, then one might expect improved whole body glucose tolerance by virtue of enhanced insulin sensitivity.

Our laboratory's recent observations indicate that oral creatine supplementation provides a valuable countermeasure of immobilization-induced reduction in muscle GLUT-4 content in healthy humans (20). Furthermore, after the subsequent 10 wk of rehabilitation training, the GLUT-4 content was still elevated by creatine supplementation (20). Finally, this study also showed that creatine supplementation in conjunction with training is able to potentiate glycogen accumulation in muscle, a finding that supported other recent studies, including rat experiments (18, 19, 25).

The present study was undertaken to further explore the effects of oral creatine supplementation on skeletal muscle GLUT-4 and glycogen content. Specifically, we evaluated whether the effect of creatine supplementation is limited to the muscle group of which the activity level is decreased/increased or whether it is also present in muscles not subjected to changes in activity level. Robinson et al. (25) showed that after a single bout of submaximal one-legged cycling exercise, muscle glycogen accumulation is enhanced by creatine supplementation in the exercised but not in the rested leg. We presently aim to expand these observations by evaluating the effect of 6 wk of one-legged heavy resistance training on muscle glycogen and GLUT-4 content. If the impact of creatine intake on GLUT-4 content is restricted to the limited group of trained muscles, then one would only expect a minor improvement in whole body glucose tolerance. We have tested this hypothesis in the present study by performing an oral glucose tolerance test (OGTT) before and after supplementation.

It is presently unknown by what mechanism creatine supplementation potentiates the increase in GLUT-4 expression due to training. The regulation of GLUT-4 biogenesis in skeletal muscle is still poorly understood (16, 32). The training-induced stimulation of GLUT-4 expression is known to occur in the recovery period after an exercise bout at both a pretranslational and translational level (15). Creatine supplementation is thought to have anabolic effects, because it is able to promote training-induced muscle hypertrophy (13, 31). Recent studies have shown that the ingestion of a protein lysate supplement before, or immediately after, a heavy resistance training session stimulates muscle hypertrophy (7, 28). Therefore, it could be hypothesized that the addition of protein and amino acids to the creatine supplement may further stimulate training-induced muscle hypertrophy and GLUT-4 protein synthesis by providing additional amino acids in the postexercise period. We have explored this hypothesis by daily supplying creatine (2.5 g) in the presence and absence of 40 g of a protein blend (whey, milk, casein, and egg albumin protein) and 6.7 g of an amino acid blend (glutamine, leucine, valine, isoleucine, methionine, and lysine) delivered immediately after each training session. The resulting hyperaminoacidemia is expected to further enhance the promoting effect on hypertrophy and GLUT-4 expression induced by training and creatine.

The present study examined the effects of creatine and creatine plus protein supplementation on the GLUT-4 and glycogen content in human skeletal muscles either in the absence or in the presence of immobilization (2 wk) and with subsequent retraining (6 wk).

	RESEARCH DESIGN AND METHODS

TOP ABSTRACT INTRODUCTION RESEARCH DESIGN AND METHODS RESULTS DISCUSSION REFERENCES

Subjects. Thirty-three healthy subjects {26 men and 7 women; age 18-30 yr [average 19.8 ± 0.5 (SE) yr]} gave their written informed consent to participate in the study. Exclusion criteria were creatine intake in the 4 mo preceding the study, vegetarianism, prior or existing renal pathology, any pathology that is contraindicative to leg immobilization and/or training, and consistent intake of any medication. The study protocol was approved by the local Ethics Committee.

Study design. In this double-blind, placebo-controlled trial, subjects were assigned to one of three supplement groups that were randomized for gender, body weight and maximal isometric torque of the knee extensor muscles (see below). After baseline measurements, a light polyester cast, extending from groin to ankle, immobilized each subject's right leg at a knee angle of ~160° for 2 wk. After removal of the cast, the same measurements as at baseline were performed. Thereafter, the right leg was retrained during a 6-wk rehabilitation program. During this retraining period, subjects came every other day to perform 3-5 series of 12 unilateral knee extensions with the right leg. The training load was gradually increased from three series at 75% of one repetition maximum (reevaluated once per week) at the start of the training period, to five series at 85% of one repetition maximum toward the end of the study. All training sessions were supervised by one of the investigators. The subjects were randomly assigned to one of three groups receiving different supplements during immobilization and retraining (see Table 1). The placebo group (Pl) received maltodextrine supplements that were made isocaloric to the supplements of the other groups. The creatine group (Cr) received creatine at a dosage of three daily doses of 5 g per day during immobilization, decreasing to a single 2.5-g daily dose during retraining. The creatine plus protein group (Cr+P) received 3 doses of 5 g creatine per day during immobilization and one dose of 2.5 g creatine, 40 g protein, and 6.7 g amino acids during retraining. All supplements were prepared to be isocaloric and similar in taste by the addition of maltodextrine and artificial flavor.

Measurements. At baseline, after 2 wk of immobilization (~1 h after removal of the cast) and after 6 wk of retraining (>48 h after the last training session), percutaneous needle biopsies were taken under local anesthesia (2-3 ml of 1% lidocaine) from the vastus lateralis muscle of both legs. Part of the muscle sample was immediately frozen in liquid nitrogen, and the remaining part was mounted in embedding medium, cooled in isopentane, and stored for later histochemical analyses. On the evening and morning before each biopsy session, subjects received a standardized dinner (~850 kcal, 47-25-28% carbohydrate-fat-protein distribution) and a light standardized breakfast (~320 kcal, 65-15-20% carbohydrate-fat-protein distribution). In the morning of the second day after the biopsy and after an overnight fast, subjects underwent a 2-h OGTT (75 g). Capillary blood samples for the measurement of glucose concentration were collected from a hyperemic earlobe at 20-min intervals, and venous blood samples for the measurements of insulin were taken from an antecubital vein at 1-h intervals. In the afternoon, isometric and dynamic maximal knee extension torques of the right leg were evaluated by using an isokinetic dynamometer (29), and the left leg was evaluated 10 min after completion of the protocol with the right leg. In sitting position, the subjects performed three voluntary maximal isometric contractions (3 s) at a knee angle of 110°, and the best of the three attempts was taken as peak torque. Thereafter, they performed a bout of 30 maximal dynamic knee extensions at a constant velocity of 180°/s in a constant range of 90°. After 2 min of recovery, the same bout of 30 extensions was repeated. A recovery index was calculated as the average torque of the first five contractions of the second bout, relative to the average of the last five contractions of the first bout.

Biochemical analyses. Blood samples were analyzed for glucose concentration by using an automated glucose analyzer (model 2300 STAT, Yellow Springs Instruments, Yellow Springs, OH), and for insulin concentration by using a radioimmunoassay method (insulin RIA kit, Amersham Pharmacia Biotech, Uppsala, Sweden). The  insulin concentration is the absolute insulin concentration (in µU/ml) at 60 min of the OGTT minus the fasting insulin concentration. The suprabasal area under the glucose curve was calculated by using the trapezoidal rule. Muscle samples were freeze-dried and dissected free of blood, fat, and connective tissue. Glycogen, ATP, free creatine, and phosphocreatine were determined by standard enzymatic fluorometric assays (10). Muscle free creatine and phosphocreatine values for each individual were corrected for the mean ATP content over the three time points. For the subjects from whom sufficient muscle tissue was available in all six biopsies (9 of 11 per group), GLUT-4 measurements was performed with Western blot analysis. In brief, 20-30 mg of wet muscle were homogenized (Polytron) on ice in a buffer (pH 7.4) with the following composition: 250 mmol/l sucrose, 20 mmol/l NaHCO₃, 1 mmol/l phenylmethylsulfonyl fluoride, 0.5% nonionic surfactant (Triton X-100), and a protease inhibitor cocktail tablet (Complete, Roche). The homogenate was then incubated under gentle agitation at 4°C for 30 min and spun for 15 min at 13,000 g (4°C). The protein concentration of the supernatant was determined by using a BCA protein assay kit (Pierce). Then, 45 µg of supernatant protein were separated by SDS-PAGE with 10% Tris-glycine polyacrylamide gels. After the proteins were transferred to a polyvinylidene fluoride membrane, the membrane was blocked at room temperature for 1 h in TBS-T (Tris-buffered saline containing 0.05% Tween 20, pH 7.4) comprising 5% skimmed milk powder. Blocked membranes were incubated overnight at 4°C in TBS-T comprising 1% skimmed milk powder and a specific rabbit polyclonal antibody against 13 COOH-terminal amino acids of GLUT-4 (AB1346, Chemicon International, Temecula, CA). Finally, GLUT-4 was visualized with an alkaline phosphatase-conjugated goat anti-rabbit secondary antibody (DAKO, Copenhagen, Denmark) by fluorescence imaging by using a Kodak Image Station 440 CF, and band density was calculated by using Kodak 1D image analysis software. To correct for intergel differences, band intensity was expressed relative to a standard that was run together with the samples. For the histochemical analyses on baseline samples only, serial sections (10 µm) were cut with a freezing microtome and stained for myofibrillar ATPase to allow identification of the different fiber types (2). Relative fiber type and area distribution and mean fiber area were measured by a video camera-connected microscope and computer equipped with specially developed software (Tema, Scanbeam, Hadsund, Denmark).

Statistics. Treatment effects during immobilization and retraining were evaluated by a two-way (repeated measures) ANOVA, which was covariate adjusted for the baseline or postimmobilization values (Statistica, Statsoft, Tulsa, OK). Time effects were analyzed with two-way ANOVA without covariate adjustment or with one-way ANOVA (repeated measures) to compare effects of immobilization and retraining within each group. Post hoc Tukey's tests were performed to locate the pairwise differences. Data are presented as means ± SE, and P < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION RESEARCH DESIGN AND METHODS RESULTS DISCUSSION REFERENCES

Muscle GLUT-4. The protein expression of the glucose transporter (GLUT-4) is shown in Fig. 1. In the right leg, compared with baseline, the immobilization decreased GLUT-4 in Pl and Cr (P < 0.05) but not in Cr+P (Fig. 1A). However, these changes were not significantly different between the groups. After retraining, GLUT-4 content was not different from baseline level in Pl. Conversely, in Cr and Cr+P, increases of GLUT-4 expression at the end of the retraining period were greater than in Pl (P < 0.05). Thus, after retraining, GLUT-4 protein expression in Cr and Cr+P was higher than postimmobilization (P < 0.05). In the left control leg, GLUT-4 content was constant throughout the study in all groups (Fig. 1B).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   GLUT-4 expression. Protein expression of the glucose transporter GLUT-4 in muscle samples of the right (experimental leg; A) and left (control leg; B) leg of subjects receiving placebo (Pl group), creatine (Cr group), or creatine plus protein (Cr+P) supplements. Values are means ± SE of 9 observations per group. Post-Immo, postimmbolization; arb, arbitrary. # Difference compared with baseline with 1-way ANOVA, P < 0.05. § Difference compared with Post-Immo with 1-way ANOVA, P < 0.05. * Significant group-treatment interaction with 2-way analysis of covariance shows post hoc difference compared with Pl, P < 0.05 (Tukey's test).

Muscle glycogen. During the immobilization period, the muscle glycogen content remained unchanged (Fig. 2A). After retraining, the glycogen content of the right leg was similar to baseline in Pl, whereas in Cr and Cr+P, glycogen content was supercompensated, reaching values of nearly 500 mmol/kg dry wt, which were ~35% higher than in P (P < 0.05). The glycogen content of the left leg remained unaltered throughout the experimental period in any group (Fig. 2B).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Muscle glycogen content. Glycogen concentrations in muscle samples of the right (experimental leg; A) and left (control leg; B) leg of subjects in Pl, Cr, and Cr+P groups. Values are means ± SE of 11 observations per group. dw, Dry weight. § Difference compared with Post-Immo with 1-way ANOVA, P < 0.05. * Significant group-treatment interaction with 2-way analysis of covariance shows post hoc difference compared with Pl, P < 0.05 (Tukey's test).

Muscle creatine content. As shown in Table 2, muscle total creatine content remained in the range of 120-130 mmol/kg dry wt in the Pl group in both legs throughout the study. Compared with baseline, total creatine in Cr and Cr+P tended to increase in both legs during the immobilization period (P < 0.05 for left leg in Cr; P < 0.10 for the other conditions), yet these changes were not significantly different from placebo. In the creatine-supplemented groups (Cr and Cr+P), total creatine further increased (~30 mmol/kg dry wt) higher than baseline during the retraining period in the trained leg (P < 0.05 vs. baseline in Cr and Cr+P) but not in the control (left) leg. The increases in total creatine content resulted from elevations in both the free creatine and phosphocreatine content. Compared with Pl, phosphocreatine content of Cr+P was significantly increased in both legs at the end of the study (treatment effect, P < 0.05).

OGTT. The results of the OGTT are shown in Table 3. The area under the glucose curve is unaltered over time in Pl and Cr. However, in Cr+P, the area under the glucose curve at the end of the intervention period was ~35% lower compared with Pl (P < 0.05), and ~25% lower compared with baseline (P < 0.05), which indicates an improved glucose tolerance. The plasma insulin levels were not affected by the immobilization and retraining interventions in any group at 60 min (Table 3) and 120 min (data not shown) during the OGTT.

Muscle fiber type composition. Histochemical analyses were performed on biopsy samples obtained at baseline. The relative fiber type distribution in the vastus lateralis of the right leg was similar between groups. The relative number of type I, type IIa, and type IIb/x fibers, respectively, was 50 ± 5, 28 ± 3, and 21 ± 4% for Pl; 53 ± 3, 32 ± 2, and 15 ± 2% for Cr; and 54 ± 3, 31 ± 2 and 15 ± 2% for Cr+P. Similarly, the relative fiber area distribution did not differ between groups (data not shown). The mean fiber area of type I, type IIa, and type IIb/x fibers, respectively, was 4,940 ± 530, 5,396 ± 851, and 4,650 ± 875 µm² for Pl; 4,837 ± 694, 5,636 ± 688, and 4,996 ± 811 µm² for Cr; and 4,923 ± 393, 5,584 ± 472, and 4,776 ± 530 µm² for Cr+P.

Muscle force. In the right leg, the maximal torque decreased by ~30% during immobilization and recovered to values above baseline toward the end of the 6-wk retraining period (Table 4). These changes in maximal torque were similar between groups throughout the study. In the left leg, maximal torque was not affected by treatment or time. In the right leg, the recovery index was not affected by immobilization, but it increased with training. Compared with baseline, in Cr and Cr+P, but not in Pl, the recovery index was ~15% higher after retraining (P < 0.05; Table 4). The recovery index was unaffected in the left control leg.

Body weight. Body weight was 69.7 ± 2.4, 67.2 ± 2.1, and 68.4 ± 2.4 kg at baseline in Pl, Cr, and Cr+P, respectively, and increased to 70.1 ± 2.3, 68.5 ± 2.1, and 69.9 ± 2.4 kg (P < 0.05 vs. baseline for Cr and Cr+P) after 2 wk of immobilization. Body weight increased further during retraining to 71.1 ± 2.2, 69.3 ± 2.2, and 71.2 ± 2.5 kg (P < 0.05 vs. baseline for all groups), respectively. No treatment effects were observed for body weight.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESEARCH DESIGN AND METHODS RESULTS DISCUSSION REFERENCES

Our laboratory previously demonstrated that oral creatine supplementation can increase skeletal muscle GLUT-4 and glycogen content in healthy humans involved in an exercise training program after leg immobilization (20). Our present study confirms this finding, yet, in addition, we here,for the first time, clearly show that creatine supplementation can increase muscle GLUT-4 and glycogen content only on the condition that it is administered in conjunction with increases in muscle activity level. Evidence for such concerted action of creatine intake and muscle contractile activity comes from the observation that in the contralateral (nonimmobilized and nontrained) leg, glycogen and GLUT-4 content was not influenced by creatine supplementation. Thus creatine supplementation per se has no effect on either muscle GLUT-4 expression or glycogen content, but clearly it can enhance the beneficial impact of training on it. These findings are in line with the study by Robinson et al. (25), who found a tendency (P = 0.06) for increased muscle glycogen accumulation after creatine supplementation in knee extensor muscles that were previously exercised but not in the nonexercised contralateral legs.

A major aim of the present study was to evaluate whether the addition of protein and amino acids would stimulate the beneficial effect of creatine supplementation on muscle glycogen and GLUT-4 content during retraining after immobilization. The results show that the effects of combined creatine and protein supplementation to increase muscle GLUT-4 and glycogen content in the retrained leg were similar to the effects of creatine alone. Accordingly, combined creatine and protein supplementation did not cause changes in either GLUT-4 or glycogen in the contralateral leg. There is considerable evidence in the literature indicating that oral protein/amino acid supplementation of similar composition, at an even lower dose than in the present study, elicits hyperaminoacidemia and elevated muscle amino acid uptake after resistance exercise (22, 30). Thus increased availability of amino acids after a training bout does not appear to stimulate net GLUT-4 protein synthesis.

Combined creatine and protein supplementation (Cr+P group) during the 6-wk rehabilitation period resulted in a substantial reduction of the area under the glucose curve during a 2-h OGTT, which indicates improved glucose tolerance in this condition. This could either be related to improved pancreatic insulin secretion or to increased peripheral insulin sensitivity. There is a single study to suggest that creatine supplementation might stimulate pancreatic insulin release (17). Indeed, supraphysiological concentrations of guanidinoacetate compounds have been found in vitro to stimulate pancreatic insulin release (1). Although some studies indicate that acute or chronic oral creatine intake does not stimulate insulin secretion (8, 19), a recent study by Rooney et al. (26) provides evidence for elevated insulin secretion after 4-8 wk of creatine supplementation in rats. The present study on the other hand suggests that insulin concentrations during the OGTT were independent of creatine supplementation. Thus the improved glucose tolerance seen after creatine plus protein supplementation is likely due to improved peripheral insulin sensitivity rather than to facilitated insulin secretion. Because, however, the trained knee extensor muscle group represents only a small part of total muscle mass, it is unlikely that improved glucose tolerance in the Cr+P group is caused by the elevation of GLUT-4 expression in only the experimental leg. However, creatine plus protein supplementation did not increase GLUT-4 expression in the contralateral control leg. In fact, there is substantial evidence to indicate that changes in insulin sensitivity can occur independent of changes in skeletal muscle GLUT-4 expression (21). Further experiments are needed to evaluate whether either improved peripheral or hepatic insulin sensitivity can explain the improved oral glucose tolerance in healthy subjects caused by combined creatine and protein supplementation.

In the present study, creatine supplementation promoted the training-induced increase but did not prevent the immobilization-induced decrease in muscle GLUT-4 content. This is in apparent contrast with our laboratory's previously published observations (20), in which we also observed an effect during immobilization. Part of the explanation may lie in the fact that creatine supplementation did not cause significant elevation of the muscle total creatine content during immobilization in Cr and Cr+P (0.05 < P < 0.10 for right legs in Cr and Cr+P), whereas in our laboratory's previous study the elevation was significant compared with baseline and placebo (20). This therefore suggests that a marked elevation of muscle total creatine content is a prerequisite to other effects of creatine supplementation. However, during immobilization, it is more difficult to achieve a large elevation of muscle total creatine content, because muscular activity is known to promote creatine uptake in muscle (25), as is further illustrated by our posttraining findings.

It is well established that the different human muscle fiber types display significant differences with regard to the GLUT-4, creatine, and glycogen content and, more importantly, with regard to their responsiveness to various stimuli, such as training (3-5). Therefore, we measured fiber type distribution of the vastus lateralis muscle at baseline. Relative distribution of type I, IIa, and IIx/b fibers was similar between the experimental groups, which excludes muscle fiber composition to be a confounding factor in the interpretation of the present results.

Muscle disuse and retraining cause significant atrophy and hypertrophy, respectively. The results on the maximal isometric knee extension torque and maximal dynamic knee extension power (data not shown) show a significant reduction (30% vs. baseline) during immobilization and full recovery (+5 to +15% vs. baseline) during retraining. The lack of difference between the experimental groups is partly in contrast with our laboratory's previous findings (13), which indicated that dynamical power, but not isometric torque, was enhanced by creatine supplementation during rehabilitation training. This discrepancy could relate to the difference in initial muscle creatine content, which is known to be an important determinant for the efficacy of creatine supplementation (11, 12). Additionally, the supplement dosage and training intensity and duration differed between the present and our laboratory's previous study design (13). Facilitated muscle phosphocreatine resynthesis is believed to be an important mechanism underlying enhanced muscle power output during intermittent, high-intensity muscle contractions (9). We did not measure muscle phosphocreatine contents during recovery from our maximal intermittent exercise test for the knee extensor muscles. However, the recovery of torque during the 2-min rest interval after 30 maximal concentric contractions (indicated in Table 4 as the recovery index) was markedly increased (P < 0.05) by the training intervention in the two creatine-supplemented groups (Cr and Cr+P) but not in the Pl group. Thus, although the present data on muscle force are not entirely in conformity with our laboratory's previous study, these data on recovery index confirm that oral creatine supplementation specifically improves some training-induced effects on muscle contractile function, as previously suggested (13).

In conclusion, the present data for the first time indicate that combined creatine and protein supplementation has the potential to improve oral glucose tolerance in healthy subjects, a finding that is probably not related to changes in muscle GLUT-4 expression level. Furthermore, our observations show that oral creatine supplementation increases muscle GLUT-4 and glycogen content in human skeletal muscles, exclusively when they undergo increases in activity pattern, such as during resistance training. The addition of protein supplement and the resulting hyperaminoacidemia immediately after the training bouts do not further enhance these effects of creatine during the retraining period after immobilization.