INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE PRESENT UNDERSTANDING of skeletal muscle metabolism in children is based almost exclusively on a small number of muscle biopsy studies conducted more than 25 years ago in small groups of 11- to 15-yr-old boys (7, 9, 10). This research reported a possible relationship between anaerobic enzyme activity and maturation, initiating the concept of limited anaerobic capacity in children. These findings have been so regularly cited that a limited anaerobic metabolism in children has become an accepted paradigm in the pediatric exercise literature. However, others, using biopsy methods, have been unable to demonstrate such a relationship between maturation and glycolytic enzyme activities (3, 13, 14). The ethical limitations of biopsy techniques have made it difficult to study muscle metabolism in children. Consequently, the development of metabolic pathways in children and the consequent metabolic responses during exercise have not been well documented. Phosphorus nuclear magnetic resonance spectroscopy (³¹P-MRS) is a powerful, noninvasive method of evaluating selected aspects of muscle metabolism in vivo in adults (4, 15, 21, 22, 32) and children (19, 28, 31) and offers the opportunity to address some of these questions.

Zanconato et al. (31) reported evidence of greater glycolytic metabolism during graded exercise to exhaustion in adult (men and women, age range 20-42 yr) calf muscle compared with that of prepubescent boys and girls (age range 7-10 yr). Unfortunately, information regarding maturity status of the children, training status of all subjects, and their motivation to exercise to exhaustion was not described, which tends to limit the interpretation of the data with respect to the question of maturity-related energy metabolism. Taylor et al. (28) also used a graded exercise protocol to evaluate metabolic responses in the calf muscle of children (age 6-12 yr), younger adults (age 20-29 yr), and older adults (age 70-83 yr). In contrast with the younger adults, the children demonstrated higher oxidative capacity, higher pH and ADP levels during exercise, and faster metabolic recovery after exercise. These results could suggest that, at the same relative intensity of exercise, the children relied less on glycolysis and more on oxidative metabolism to meet the energy demand.

Kuno et al. (19) used ³¹P-MRS to compare bioenergetics in the quadriceps muscle of trained and untrained adolescent boys with that in untrained adult controls. After the subjects performed a graded exercise protocol, the authors found significantly lower intramuscular pH [intracellular pH (pH_i)] values in the adults compared with that in either group of adolescents. Maturation status was not reported, and, considering the chronological ages (range 12-17 yr) of the boys, it would be reasonable to suspect significant variability in maturation status. There were no differences in the metabolic responses between the trained and untrained adolescents. These findings are interesting considering the significant training-induced metabolic adaptations in adolescent boys reported by Eriksson et al. (7, 8).

Whereas previous research has made comparisons between children and adults, the influence of maturation during the pubertal period on muscle metabolism remains unknown. The purpose of this study was to describe the bioenergetics of skeletal muscle in physically active prepubertal and pubertal girls during rest and submaximal and supramaximal exercise.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. After institutional research ethics board approval, female volunteers (n = 18), age 9-16 yr, with a minimum of 1 yr of training, were recruited from a local competitive synchronized swimming club. Written informed consent was obtained from subjects and parents after a detailed description of the purpose and procedures of the study.

Exercise procedures. Subjects were familiarized with the plantar flexion exercise protocol 5-10 days before the ³¹P-MRS data collection. At this time, subjects completed a graded plantar flexion maximal work capacity (MWC) test with the right calf (4). This test, conducted outside the MRS unit, involved 30 repetitions/min through a constant range of motion at the ankle while the rest of the leg and thigh were fixed in position with restraining straps. The foot pedal of the calf ergometer was connected via a simple system of rope and pulleys to a weighted basket. Each repetition required lifting the basket against gravity. Resistance was increased systematically every minute until failure. Subjects were actively encouraged to continue for as long as possible. The work done during the final exercise stage completed was designated as 100% MWC. After a short recovery, subjects practiced the MRS exercise protocol using their left leg to familiarize themselves with the exercise challenge they would encounter in the actual experiment.

NMR data collection. ³¹P-MRS and magnetic resonance imaging were conducted in a 3-T Magnex magnet interfaced with a Surrey Medical Imaging Systems console and operating system. Subjects were in a supine position within the magnet, and resting images of the right calf were taken before the exercise protocol was begun. A 25-cm-diameter leg birdcage resonator and multislice gradient echo imaging sequence were used to acquire a series of magnetic resonance images [echo time = 20 ms, repetition time (TR) = 1 s, and 5 adjacent slices with slice thickness = 5 mm]. Initially, a sagittal image was acquired to register the subsequent transverse series to an external reference water sphere. The water sphere was positioned at the level of the region of maximal calf girth, measured by anthropometry before the placement of the subject into the magnet. Each image was subsequently examined to find the greatest CSA for the two heads of gastrocnemius.

NMR data analysis. Gastrocnemius muscle CSA was estimated by using the stereology module of the Analyze image analysis package (Mayo Foundation, Biomedical Imaging Resource, Rochester, MN). In brief, a grid with spacing of 5 mm along the horizontal and vertical axes was superimposed on the transverse image of the leg. Intersecting grid lines, located over the muscle region for which the CSA was to be estimated, were counted. Because the maximum number of "hits" of the grid corresponds to the field of view (FOV) selected in the image acquisition parameters (XX × YY mm²), the selected muscle CSA was estimated as a direct ratio as follows
CSA of the lateral and medial gastrocnemius muscle heads were calculated separately from the magnetic resonance images and then combined to provide the total muscle areas shown in Table 1.

Statistical analysis. All data are reported as group means ± SD. Differences between the two maturity groups were tested by using one-way ANOVA for anthropometric measures, gastrocnemius muscle CSA, and metabolic parameters at rest and for each 30-s period of the exercise protocol. Pairwise interaction contrasts were used to identify the location of interaction effects. In this case, two-way ANOVA with repeated measures on one factor (time) was used to test for interaction effects over sequential time points (e.g., between 5.0 and 5.5 min) in the exercise protocol. Significance was set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Physical characteristics. The subjects were separated into two groups according to maturity status. The Pub group (n = 9) was significantly older, taller, and heavier, and completed more absolute work during the MWC test than did the Pre group (n = 9). The physical characteristics of the two groups are shown in Table 1.

Exercise performance. No statistically significant difference was found in the submaximal (40% MWC) or supramaximal (140% MWC) exercise loads between the groups when corrected for body mass or for gastrocnemius muscle CSA. Therefore, whereas the absolute work completed by the Pub subjects was greater, the relative amount of work was the same for all subjects.

pH_i. No statistically significant difference was found in the pH_i values between the two maturity groups at any point throughout the NMR protocol. End-exercise pH_i was 6.66 ± 0.11 and 6.76 ± 0.17 for Pub and Pre, respectively (Fig. 1).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Mean ± SD of intracellular pH (pHi) at rest and during submaximal [40% maximal work capacity (MWC)] and supramaximal (140% MWC) exercise in prepubertal (Pre: n = 9) and pubertal girls (Pub: n = 9).

P_i/PCr. Fig. 2 displays the change in the ratio of P_i to PCr (P_i/PCr) from resting through the exercise protocols. P_i/PCr did not differ between the two groups at rest (Pre: 0.043 ± 0.01; Pub: 0.083 ± 0.04) or during either of the exercise protocols. During the submaximal exercise, both groups experienced a slight, but consistent, increase in the P_i/PCr. Supramaximal exercise elicited a more distinct increase in P_i/PCr in both groups, with the highest values observed at end exercise (Pre: 1.31 ± 0.88; Pub: 2.18 ± 1.00). The mean P_i/PCr values were not significantly different; however, ANOVA revealed a significant time-by-group interaction effect during the supramaximal exercise challenge. Pairwise interaction contrasts were employed to locate a significant interaction effect (1 df; F = 5.377; P = 0.034) between 6.0 and 6.5 min of the protocol. There was no significant interaction for the subsequent time intervals (6.5 and 7.0 min) of the protocol (1 df; F = 2.47; P = 0.14).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Mean ± SD of ratio between Pi and phosphocreatine (PCr) (Pi/PCr) at rest and during submaximal (40% MWC) and supramaximal (140% MWC) exercise in Pre (n = 9) and Pub girls (n = 9).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the pediatric literature, it is commonly acknowledged that anaerobic metabolism is related to physical maturity and thus is limited in children relative to adults. Lower peak blood lactate concentrations (1, 7, 10, 20) and lower anaerobic power in children compared with adults (12, 17, 27) have been used to support the contention that maturation influences anaerobic development. There are a number of methodological explanations for the low blood lactates often reported in the literature, including the intensity of exercise used to induce peak lactate values, motivation of the subjects, measurement techniques, and muscle size. In addition, whereas a relationship between blood lactate and testosterone levels has been demonstrated in animal models (2, 6, 18), there is no evidence of increased maximal blood lactate being influenced by the stage of sexual maturation in humans (11, 23, 24, 26, 30).

Nevertheless, blood lactate concentration is a weak predictor of glycolytic activity in working muscle. Blood lactate concentration is at best a reflection of the balance between rate of production and removal. Both production and removal are mediated by a wide variety of variables that can rarely be accounted for, considering that the metabolite is of necessity being sampled from another tissue distal from the site of production. Muscle lactate may provide a closer approximation of the metabolic environment within the tissue that produced it, but the concentration is still influenced by the balance of production and removal. NMR allows the measurement of pH_i from the tissue being scanned. Measurements of in vivo muscle pH can provide an accurate indication of the metabolic environment in working muscle, yet the value must still be considered in light of the balance between H⁺ production and removal at the time of measurement. The pH_i calculated from the chemical shift in the phosphorus spectra represents an indication of the results of glycolysis, not necessarily a measure of glycolytic activity.

The NMR studies in children (19, 28, 32) tend to support the concept of maturity-related metabolic responses during heavy exercise. These authors reported evidence of less glycolytic activity in children compared with adults. However, these studies were not designed to address the significance of maturational changes associated with puberty. The intention of the present study was to investigate the pubertal influences on metabolic characteristics of skeletal muscle.

Our results demonstrate lower pH_i levels in calf muscle during heavy exercise than have previously been reported for prepubescent children. Zanconato et al. (31) used an exercise protocol similar to our MWC test and reported an average end-exercise pH_i of 6.93, which represented a net decrease of 0.11 pH units. The mean pH_i at end exercise for the prepubescent girls in the present study was 6.76 (Fig. 1), which represents a decline of 0.33 pH unit. Similarly, the increase in P_i/PCr for the Pre group at end exercise (Fig. 2) is considerably more dramatic than might be expected based on the observations of Zanconato et al. In that experiment, the ratio for the children at end exercise was 0.54 ± 0.12, which contrasts with the present results of 1.31 ± 0.88.

These differences may be partly explained by the exercise protocols utilized. Considerably greater glycolytic involvement should result from the supramaximal exercise challenge used in the present study, compared with the graded exercise protocols previously employed (19, 28, 31). In addition, based on the significant training-induced adaptations reported by Eriksson et al. (7, 8) in the muscles of young boys, training status should be considered as a potential influence on muscle bioenergetics during exercise. We are not aware of other reports from studies of trained prepubescent muscle during supramaximal exercise, which makes comparison with results from other studies difficult.

We found no significant differences between the two maturity groups in the metabolic responses at any point in the protocol. The end-exercise pH_i values for both Pre and Pub groups were consistent with previous findings from this laboratory using similar exercise challenges in trained and untrained male and female adults (4, 15, 16, 25). These findings do not support the concept that glycolysis is attenuated in prepubescent muscle. It has been frequently suggested (3, 7, 14) that, at similar relative exercise intensities, prepubescent children rely more on oxidative metabolism than do their adult counterparts. The exercise loads in this experiment were equal when adjusted for both body weight and gastrocnemius CSA and elicited metabolic responses of similar magnitude. This observation leads to the conclusion that muscle bioenergetics are not maturity dependent, as has been previously suggested. This conclusion should be drawn with some caution. The narrow age range in the present study was intended to span the pubertal period. The results of other studies (19, 28, 31) comparing children with adults may be affected by factors other than maturity, such as growth and training history.

Although the end-exercise pH and P_i/PCr values were not statistically significant, an interaction effect (P < 0.05) was revealed for P_i/PCr during the supramaximal exercise. This finding, in combination with the tendency for lower pH_i values for Pub subjects at the same time points, could be interpreted as evidence for maturation-based differences in muscle metabolism. That is, the Pub subjects may have relied more on glycolysis, and the Pre subjects may have relied more on oxidative metabolism as described above and suggested by the results of Taylor et al. (28).

However, we believe that a more plausible explanation is available based on morphological differences between the two groups. In brief, the Pub subjects had, on average, considerably more gastrocnemius muscle than did the Pre subjects, a fact that may have a significant impact on the interpretation of the unlocalized MRS data.

Several factors must be considered regarding the signal source, which influences the interpretation of MRS spectra from calf muscle. These include the potential differences in the thickness and CSA of gastrocnemius muscle, the well-known differences in fiber composition of soleus and gastrocnemius, and the relative importance of the gastrocnemius muscle to power production during intense exercise. When a surface coil is used for both transmission and signal reception, the ³¹P signal is acquired most efficiently closest to the coil. The coil sensitivity, in the "receive" mode, falls proportional to the square of the distance from the coil. Therefore, the majority of tissue contributing to the ³¹P spectrum will come from one coil radius depth (4 cm in this study), although there will be an ever smaller contribution from deeper lying tissues. Consequently, the unlocalized MRS signal will contain signal from both gastrocnemius and soleus muscles in relative proportions to their thickness and distance from the transceiver coil.

It is possible to employ techniques that localize the MRS signal to a specific region of interest. Zhu et al. (32) reported significantly different metabolic responses from the soleus muscle compared with either head of the gastrocnemius muscle during calf exercise. This is not surprising, because it is well accepted that the soleus is composed mainly of type I and gastrocnemius mainly of type II muscle fibers. It is also known that gastrocnemius muscle is the primary contributor of muscle power during intense plantar flexion exercise (29).

The technique described by Zhu et al. (32) was attempted in the present study but was unsuccessful due to the limited size of the prepubescent calf musculature. The thickness of gastrocnemius is of critical importance in this technique. From magnetic resonance images of the leg, it was determined that the mean gastrocnemius muscle thickness in the Pub group was ~44% greater than that in the Pre group (13 vs. 9 mm). Logically, if the greater proportion of MRS signal is obtained from gastrocnemius muscle rather than from the deeper soleus muscle, then the metabolic responses represented by the mixed signal will reflect the type II gastrocnemius fibers, which favor rapid glycolysis under the supramaximal exercise conditions. The tendency toward lower pH_i and higher P_i/PCr in the Pub group may logically be explained by a greater proportion of the signal coming from larger gastrocnemius muscles in these subjects.

The importance of these morphological differences do not appear to have been fully considered by others using ³¹P-MRS to interrogate calf muscle in children (28, 31). Unless signal localization techniques to specific muscles are utilized, it is suggested that comparisons between adults and children should be undertaken with caution. Failure to consider the impact of the potential influence of both muscle size and fiber type differences between soleus and gastrocnemius when comparing adults and children may lead to spurious conclusions.

Nevertheless, such arguments do not preclude the possibility that there may be some maturation-related differences in muscle metabolism between prepubertal and pubertal children and/or adults. However, we suggest that at least part of any apparent difference in metabolic responses between our subject groups can be explained by morphological differences, which then reduces the potential significance of any differential metabolic factors.

Because little is known about acute responses or chronic adaptations of skeletal muscle to heavy exercise in children, it is very important to continue investigating the bioenergetics of skeletal muscle in trained and untrained children while avoiding, or at least considering, the morphological limitations described above. One approach might be to focus on subjects of similar chronological age and maturation status, because then the muscles under study would be approximately the same size. The developmental and physical similarities should eliminate the morphological concerns detailed above while allowing investigation of the effects of exercise and training. Such a direction would seem to be a logical step in the study of muscle metabolism in children.