Energy Cost and Energy Sources of Ball Routine in Rhythmic Gymnasts

Training and Testing 205 Energy Cost and Energy Sources of Ball Routine in Rhythmic Gymnasts L. Guidetti 1, C. Baldari 1, L. Capranica 1, C. Persichini 1, F. Figura 2 1 University Institute of Motor Sciences (IUSM) of Rome, Italy 2 Dept. of Human Physiology, University of Rome La Sapienza, Italy Guidetti L, Baldari C, Capranica L, Persichini C, Figura F. Energy Cost and Energy Sources of Ball Routine in Rhythmic Gymnasts. Int J Sports Med 2000; 21: 205 209 Accepted after revision: October 8, 1999 The energy cost and the different energy sources of competitive rhythmic ball-routines were assessed in nine young elite rhythmic gymnasts (13 16 years of age). The overall energy requirement of ball exercise (VO 2 eq) was obtained by adding the amount of VO 2 during exercise above resting (VO 2 ex) to the VO 2 up to the fast component of recovery (VO 2 al) and to the energy equivalent of peak blood lactate accumulation (VO 2 la ) of recovery. The lactate (La) curve up to 25 minutes of recovery always showed La peaks at 3 min after exercise (4.0 ± 0.4 mmol/l) and values still above rest at 25 min (2.3 ± 0.5 mmol/l). Although ball routines were short in duration (90 s), the metabolic power requirement was 1.1 times higher than the subject's V O 2 max. The energy cost (VO 2 eq) amounted to 81 ± 5ml kg 1. Higher fractions of VO 2 eq were VO 2 ex 49 % and VO 2 al 42 %, while the remaining 9 % was due to VO 2 la. The aerobic source resulted in the most utilized system due to the subjects' high V O 2 max (52.7 ml kg 1 min 1 ) and anaerobic threshold (LT = 84.4 %). The HR and the peak lactate values of ball routine were similar to the values recorded at LT step during treadmill test. Although the HR and V O 2 values were significantly correlated (during preexercise, exercise, and fast recovery), the exercise intensity derived from HR Kanvonen method during ball routine did not correspond to the % of V O 2 max. Key words: Rhythmic gymnastics, aerobic power, blood lactate concentration, anaerobic threshold, energy sources. Rhythmic gymnastics increased in popularity after the inclusion in the 1984 Olympic games in Los Angeles. The competition in Rhythmic Sportive Gymnastics includes five hand implements recognized by the International Gymnastics Federation: ball, hoop, clubs, ribbon, and rope. The athlete has to perform a physically demanding routine of 1 1 min 30 s with each of the 5 implements. The evaluation of the routine is based on precision, grace, originality, coordination to music and must contain some technical difficulties scored as A and B level. In order to avoid penalty scores, the routine has to be performed as a continuous movement flow, and even a balanced posture on one foot has to be combined with the movements of other body parts and of the implement [7]. Therefore, even if with some approximation, rhythmic routines have been considered a continuous exercise involving different types of contraction by different muscle groups. Although several studies have been published describing the physiological characteristics of elite rhythmic gymnasts [1, 2, 4, 9], few studies have been carried out on the contribution of the energy systems in rhythmic gymnastics performances [1]. Alexander et al. [1] estimated the energy expenditure from heart rate and blood lactate data. From high heart rate values during routines and frequent low lactate levels measured 5 min after the performance, the authors inferred that the most taxed energy systems in this sport activity were the aerobic and anaerobic alactic systems. They also claimed the need for further studies to be conducted to determine the direct oxygen consumption (V O 2 ) measurements during routines and to evaluate the anaerobic threshold (LT) of rhythmic gymnastic athletes, which could be a critical factor in performance. Aim of this study was to investigate, in nine young elite rhythmic gymnasts, the energy requirement of a representative rhythmic exercise, i. e. ball-routine [1], and to analyze the contribution from the different energy sources in order to relate the energy sources with the subject's V O 2 max and LT. Introduction Int J Sports Med 2000; 21: 205 209 Georg Thieme Verlag Stuttgart New York ISSN 0172-4622 Methods Subjects Nine Italian elite rhythmic gymnasts (ranging in age 14 16 yr) participated in this study. The athletes were randomly selected from the Italian Gymnastics Federation rankings, and their participation in the study was requested through their coaches. Parental permission and informed consent were obtained from all participants. This study was approved by the Ethics

206 Int J Sports Med 2000; 21 Guidetti L et al and Research Committee of the University Institute of Motor Science of Rome. The athletes had had at least 7 years of training for at least 3 hrs/day, at least four days/week, with a mean training time of 20 hours weekly. Experimental procedure All subjects underwent two testing sessions. During the first session V O 2 max and LT were assessed. In the second session the energy cost of their ball-routine was evaluated. To avoid the effects of fatigue and muscle soreness on performance, the two experimental sessions were performed one week apart. First testing session To determine her V O 2 max and LT, each athlete performed a maximal continuous treadmill test. V O 2, ventilation (V E), and heart rate (HR) were recorded as averaged values every 30 seconds during the test by a telemetric open-circuit oxygen uptake measurement system (K2 COSMED). The K2 COSMED flow meter was calibrated with a 3-L syringe, and the O 2 analyzer was calibrated with known gas mixtures (16 % and 20.9 % O 2 ). The treadmill test consisted of a 3 min walking warm-up at 6 km/h with a 0% of slope, followed by 3 min step running with a speed increment of 2 km/h each step up to the workload subsequent to LT. Then a 2 % slope increment was given each minute. When the V O 2 max was obtained, an active recovery of 1 min (walking at 5 km/h with a 0 % slope) was given. The V O 2 max was identified at the occurrence of a plateau or an increase < 1 ml kg 1 min 1 of V O 2 despite further increases in the workload [13]. The LT was defined as the V O 2 corresponding to the workload at which the blood lactate concentration rose abruptly during graded exercise. The attainment of LT was confirmed by a further increase of lactate at the subsequent workload. The average of the last minute of the corresponding work load was used to determine the oxygen consumption LT. Due to the delayed diffusion of lactate in blood [10], the measured lactate value was assigned to the workload immediately prior to that of the measurement. Determination of blood lactate was carried out immediately after the collection of untreated capillary blood from a fingertip (YSI lactate analyser model 23L USA). Blood lactate measurements were performed before exercise, at each step during exercise up to the workload subsequent to LT attainment, and at 3, 6, and 10 minutes of the recovery phase. Second testing session During the second testing session all subjects performed their competitive 90 s rhythmic ball-routines. Ball-routine can be considered representative of other rhythmic routines as indicated by the average heart rate [1]. Each competitive ball-routine included at least 6 A and 6 B difficulties [7] (Table 1). Before, during, and after the ball routines heart rate (HR), ventilation (V E), and oxygen consumption (V O 2 ) were recorded as averaged values every 15 seconds by a telemetric oxygen uptake measurements system (K2 COSMED). Due to the characteristics of K2 system the V O 2 value is not a real-time value Table 1 Difficulties A A Difficulties enclosed in each rhythmic ball routine Exercises Enjambe Saut de biche Arabesque Grand ecart Balance en passé Pivot Souplesse Tumblings and Stunts Fig. 1 Mean values of oxygen consumption (V O 2, open symbols) and heart rate (HR, solid symbols) curves before (rest and pre-exercise), during and after (fast and slow recovery) ball routine exercise. Continuous lines are regression lines of HR. Dotted lines are regression lines of V O 2. but it has to be referred to a 15 s antecedent event. This is because the real-time ventilation signal is first buffered, and then, by means of a 15 seconds delayed elaboration, it is phased with O 2 signal (10 seconds total time delay O 2 analyser). Before exercise, HR and V O 2 parameters were recorded in two different conditions: 1) at rest with the subject sitting quietly for 5 min; 2) at pre-exercise condition during which the subject reached the gymnastic mat (12 12 m) and prepared to perform her routine (Fig. 1). After exercise, HR and V O 2 were acquired during 25 minutes of sitting recovery. Determination of blood lactate was carried out immediately after the collection of untreated capillary blood from a fingertip. The measurements were performed at rest and at 3, 6, 10, 15, 20, and 25 minutes of the recovery phase. Moreover, in order to compare our data with exercises of similar duration and type as rhythmic routine [1] and artistic gymnastics [14], the exercise intensity was also calculated utilizing the Karvonen method [11]. This method corrects the exercise heart rate for the resting heart rate and compares this value with the subject's maximal heart rate reserve. Exercise HR Resting HR Percent intensity HR i % ˆ Maximum HR Resting HR 100

Energy Cost of Ball Routine in Gymnasts Int J Sports Med 2000; 21 207 Data analysis The overall energy requirement (VO 2 eq) of the exercise was obtained [8] by adding up the contributions of the three energy sources: 1. The aerobic source calculated as amount of VO 2 during exercise above resting (VO 2 ex). 2. The anaerobic alactic source calculated as the VO 2 above resting during the fast component of recovery (VO 2 al) [6]. In the recovery phase HR and V O 2 data were collected 25 min after exercise. However, since within 10 minutes of recovery V O 2 almost matched resting values and remained almost stationary during the subsequent 15 minutes, the last 15 minutes of recovery were not considered. To determine the duration of the fast and slow component of the V O 2 recovery curve, two different linear regression lines were plotted (Fig. 1): one relative to the first part of the curve (r 2 = 0.93); the other relative to the last part of the curve (r 2 = 0.81). The point of intersection between the two best fitting lines representing the point of transition between the fast and slow component of recovery occurred at 120 s. 3. The anaerobic lactic source calculated from the blood lactate accumulation (VO 2 la ) after exercise. The net increase of lactate in blood was obtained by subtracting the rest value from the peak value attained at the recovery phase. The energy equivalent of 1 mmol l 1 blood lactate increase was assumed to be 3 ml O 2 kg 1 [6]. Thus VO 2 la = la 3, where la is mmol l 1 and VO 2 la is the O 2 equivalent of the net la expressed in milliliters of oxygen per kilogram. Statistical analysis Means and standard deviations (SD) were calculated. The t-test ascertained the differences between treadmill test and ball-routine values. The relationship between HR and V O 2 was assessed by linear regression analysis, and Pearson product-moment correlation was calculated. P values less then 0.05 were considered as statistically significant. Results First testing session V O 2 max ranged from 48.1 to 57.4 ml kg 1 min 1 with a mean value of 52.7 ± 4.7 ml kg 1 min 1. HR measured at maximal effort was 201 ± 9 beats/min. LT occurred at 3.9 ± 0.3 mmol/l corresponding to the 84.4 % of V O 2 max, and the relative HR was 184 ± 17 beats/min. After treadmill test peak lactate concentrations (8.8 ± 3.6 mmol/l) were reached at the 3rd min of the recovery phase. Second testing session Mean values of rest V O 2 and HR were 3.5 ± 0.1ml kg 1 min 1 and 75 ± 5 beats/min, respectively. During ball routine the peak HR was 188 ± 5 beats/min corresponding to 93.5 % of treadmill HR max. The inter-subject average HR of the whole exercise was 177 ± 13 beats/min corresponding to 88.2% of treadmill HR max. As illustrated in Fig.1, HR increased up to the end of the exercise and decreased linearly during fast recovery (r 2 = 0.95) while during slow recovery did not show a regular tendency (r 2 = 0.08). Also V O 2 increased up to the end of the ball-routine and decreased abruptly in the fast recovery phase (r 2 = 0.93). A regular though less steep trend was maintained on the slow recovery (r 2 = 0.81) (Fig. 1). HR and V O 2 expressed in absolute values showed high and significant correlations (p < 0.001) during pre-exercise (r = 0.86), exercise (r = 0.95), and fast recovery (r = 0.98) while a low and not significant correlation was shown during slow recovery (r = 0.33). Exercise intensity quantified as % of V O 2 max and Karvonen method (HR i %) showed a similar relationship (Table 2). However, highly significant differences (p < 0.01) were found between % of V O 2 max and HR i % values of both pre-exercise and exercise conditions (Table 2). Table 2 Relationship and comparison of the % of V O 2max versus HR i % correlation coefficient p-value of paried t-test * p < 0.01 HR, V O 2, and La values during treadmill running at LT and ballroutine are presented in Table 3. Exercise intensities, expressed as HR i % and % of V O 2 max, were similar and not significantly different. Although treadmill HR values at LT were lower than ball-routine peak HR and higher than ball-routine mean HR, no significant differences were found. Moreover the ball-routine peak La during recovery was similar to treadmill La at LT. HR, V O 2, and La during treadmill running at LT and ball-rou- Table 3 tine Variables Treadmill running at LT means ± sd HR (beats/min) 184 ± 17 HR i % 85.5 ± 8.5 V O 2 as % V O 2 max 84.4 ± 0.9 Lactate (mmol/l) 3.9 ± 0.3 Ball-routine exercise % of V O 2 max versus HR i % pre-exercise exercise fast recovery slow recovery 0.86* 0.95* 0.98* 0.43 0.00001* 0.004* 0.826 0.217 Mean HR (beats/min) 177 ± 13 Peak HR (beats/min) 187 ± 5 Mean HR i % 81.3 ± 10.4 Peak HR i % 89.6 ± 6.8 Peak lactate (mmol/l) 4.0 ± 0.4 The overall energy requirement (VO 2 eq) of ball-routine exercise is described as an analysis of the contribution from the three single energy sources.

208 Int J Sports Med 2000; 21 Guidetti L et al 1. Aerobic source. The amount of O 2 consumed above resting during ball routine (VO 2 ex) was 42 ± 5ml kg 1. 2. Anaerobic alactic source. The V O 2 curve of the fast component ended at an approximate value of 12 ml kg 1 min 1. In this phase the amount of O 2 consumed above resting (VO 2 al) was 36 ± 2 ml kg 1. 3. Lactate source. In all cases the blood lactate curve determined up to 25 minutes of recovery showed the peak value at 3 min after exercise (4.0 ± 0.4 mmol/l). The last measured mean value (2.3 ± 0.5 mmol/l) was above resting value (1.4 ± 0.2). The net increase of peak lactate value ( la ) was of 2.6 ± 0.7 mmol/l. The energy equivalent of lactate accumulation in blood expressed in VO 2 la (VO 2 la = 3 la ) was 8 ± 2 ml kg 1. The contribution of the different sources to the exercise cost (VO 2 eq 86 ± 5 ml kg 1 ) is reported in Table 4. VO 2 ex and VO 2 al represented the higher fractions of VO 2 eq (49 % and 42 %, respectively) while VO 2 la was the remaining 9 %. The VO 2 eq expressed per minute (i. e. metabolic power requirement) was 57 ± 3 ml kg 1 min 1, corresponding to about 110 % of the subject V O 2 max. Table 4 Overall energy cost of ball routine exercise (VO 2 eq) in ml/kg (mean ± sd) and its partition in % of VO 2 eq among aerobic (VO 2 ex), alactic (VO 2 al), and lactate (VO 2 la ) sources Discussion ml/kg VO 2 ex 42 ± 5 49 VaO 2 al 36 ± 2 42 VO 2 la 8 ± 2 9 VO 2 eq 86 ± 5 100 % of VO 2 eq There is no data on energy cost of rhythmic routines available in the literature to compare with, although HR responses to rhythmic routines have been published [1]. Therefore, even if the main focus of this study was the energy cost and energy sources of rhythmic routine, we discuss the HR responses in order to compare our data with the literature. The high HR values we recorded during ball routines were in agreement with the data reported for exercises of similar duration in artistic and rhythmic gymnastics [1,14]. The athletes often showed high HR values at the beginning of their routine (140 160 beats/min), which may be due to both emotional stress and difficulty to maintain static starting postures [1]. Since high correlations between V O 2 and HR values measured during both pre-exercise and exercise were found, it is possible that the emotional status prior to exercise and the isometric contractions of the static starting postures could also influence the V O 2. Percentage of HR reserve is widely used in exercise prescription to estimate the percentage of V O 2 max. Thus, in the present study exercise intensity was evaluated both as V O 2 expressed in percent of subject's V O 2 max (% of V O 2 max) and as percent of subject's HR reserve (HR i %). During treadmill test no statistically significant differences were shown between the exercise intensity values calculated by means of these two methods. During ball-routine, HR i % and % of V O 2 max were closely related in the pre-exercise, exercise, and fast recovery phase. However, statistically significant differences were found between HR i % and % of V O 2 max in the pre-exercise and exercise phases, indicating that HR i % could not be used to quantify the % of V O 2 max. During fast recovery only there were no differences between the intensity values calculated by means of these two methods (Table 2). This discrepancy between HR i % and % of V O 2 max during ball-routine could be mainly due to type and duration of the exercise. Since V O 2 takes longer than HR to attain a steady state, it could be hypothesized that HR is more closely related to the intensity during the performance of an exercise of short duration. Whenever HR i % was calculated using both mean HR and peak HR, they were similar to those reported in literature for 30 70 s routines of young elite female artistic gymnasts [14]. The absolute mean and peak HR values of ball routine were also similar to those reported for other elite rhythmic gymnasts [1]. The HR and lactate data reported for the different routines with all the five implements [1] showed similar HR and La (2 3 mmol/l) values with the exception of the rope routine considered the most strenuous for the highest values of HR and La (11 mmol/l). Alexander et al. [1] inferred from only HR data during routines that much of the energy production occurred aerobically because the % of V O 2 values inferred from the HR data were likely to be below the anaerobic threshold. They also suggested a minimal contribution from the anaerobic lactic system since the majority of La peak values after routines were low. In this study the ball-routine mean and peak HR were similar to the subject's HR recorded at LT during treadmill. Also the 4.0 mmol/l of peak La measured after ball-routines indicated that the gymnasts performed their competitive exercises around their lactate threshold. The method used to quantify the energy cost and the energy sources contribution to ball-routine was described by Francescato et al. [8]. These authors studied karate exercises (lasting from 20 to 80 s) involving several muscle groups with complex movements arranged in a precise sequence. Also ball-routines were of similar duration and composed of complex movements arranged in a precise sequence. Therefore, as it occurs for many commonly practised sports (karate, soccer, tennis, basketball etc.), rhythmic routines are characterized by short spells of high intensity exercise and cannot be standardized to an meaningful extent. Because of this, there are some approximations and assumptions involved into the calculation of energy cost widely discussed by Francescato et al. [8]. They suggested that, even if some errors cannot be avoided, this method of calculation can estimate the energy cost and the percent of energy sources contribution. In the present study, analyzing the percent of energy source contribution to ball-routine, the aerobic system was found to be the most utilized (49%). Probably this is due rather to the subjects' high LT than to their fairly good V O 2 max. In fact 52.7 ml kg 1 min 1 mean V O 2 max was measured which is similar to the literature (about 50 ml kg 1 min 1 ) relative to young rhythmic gymnasts [1, 2, 4, 9], to artistic gymnasts [14], and to young female dancers [5], higher than reported for paired sedentary girls (34.5 ml kg 1 min 1 ) and lower than

Energy Cost of Ball Routine in Gymnasts Int J Sports Med 2000; 21 209 62.3 ml kg 1 min 1 of female young elite endurance athletes [3]. Besides the aerobic source, 42 % was attributable to the anaerobic alactic source (i. e. the energy phosphate splitting during exercise) and the remaining 9% to the anaerobic contribution from blood lactate accumulation. Even if these data should be regarded with some caution in view of many assumptions and approximations involved in calculation [8], they do suggest a fairly high percentage of anaerobic alactic contribution in ball routine (90 s). In karate exercise of similar duration (80 s) the anaerobic alactic source was more taxed than the aerobic source (46 % and 41 %, respectively) [8], probably due to the reported low V O 2 max (36.8 ml kg 1 min 1 ). The aerobic contribution to ball-routine (49%) fell within the range (46 % 64 %) reported for intense exercise lasting 90 s [15]. The overall metabolic power requirement of 90 seconds ballroutine is 1.1 times higher than the subject's V O 2 max. These data are lower than those (1.8) reported on 80 s karate performances by Francescato et al. [8]. No data on energy cost of gymnastic exercise are available for comparison. However, the high percentage of contribution of aerobic and anaerobic alactic sources calculated in this study seems in agreement with the hypothesis of Alexander et al. [1]. They also suggested that elite gymnasts would take advantage of a high anaerobic threshold in performing their competitive routines. Unfortunately no studies focused LT in gymnasts. Although the LT evaluated in this study is protocol specific, at LT workload a mean La value of around 4 mmol/l was found; this La value has been used to define the anaerobic threshold by some authors [10,12]. The measured LT (84 % of V O 2 max ) was comparable to that reported for endurance athletes [12] and higher than 64 % measured in young female dancers (authors' unpublished data). Therefore, as hypothesized by Alexander, maybe a high anaerobic threshold (LT) value enables a better motor control, which is important in this sport where grace, self-control, coordination, and ability to track an object are needed. Thus it is possible to conclude that: 1. The most taxed energy source during ball routine was the aerobic source. The high anaerobic threshold enables these athletes to do this high intensity type of work with relatively low levels of blood lactic acid. 2. HR and peak lactate values of ball routine were similar to those recorded at LT step during treadmill test. 3. Concerning ball routine, although HR and V O 2 were related during pre-exercise, exercise, and fast recovery phases, the exercise intensity values obtained by means of HR Karvonen method cannot be used to estimate the % V O 2 max. References 1 Alexander MJL, Boreskie SR, Law S. Heart rate response time and motion analysis of rhythmic sportive gymnastics. J Human Movement Studies 1987; 13: 473 489 2 Alexander MJL. Physiological characteristics of elite and subelite rhythmic gymnasts. J Human Movement Studies 1991; 20: 99 127 3 Burke E, Brush F. Physiological and anthropometric assessment of successful teenage female distance runners. Res Q 1979; 50: 180 187 4 Case S, Fleck S, Koehler P. Physiological and performance characteristics of the 1979 US MRG team. Internat gymnast 1980; 26: 4 5 Clarkson PM, Freedson PS, Keller BK, Carney D, Skrinar M. Maximal oxygen uptake, nutritional patterns and body composition of adolescent female ballet dancers. Res Q Exerc Sport 1985; 56: 180 184 6 di Prampero PE. Energetics of muscular exercise. Rev Physiol Biochem Pharmacol 1981; 89: 143 222 7 Federation Internationale de Gymnastique. Code de pointage gymnastique rhythmique sportive. Comité technique gymnastique rythmique sportive (F.I.G.), 1997: 23 56 8 Francescato MP, Talon T, di Prampero PE. Energy cost and energy sources in karate. Eur J Appl Physiol 1995; 71: 355 361 9 Gionet N, Babineau C, Brjant D. Physiological evaluation of elite Canadian rhythmic sportive gymnasts (RSG) (Abstract). Can J Appl Sport Sci 1986; 11: 15 10 Heck H, Mader A, Hes G, Mucke S, Muller R, Hollmann W. Justification of the 4-mmol/l lactate threshold. Int J Sports Med 1985; 6: 117 130 11 Karvonen JJ, Kentala E, Mustalo O. The effects of training on heart rate. An Med Exp Biol Fenn 1957; 33: 307 315 12 Kindermann W, Simon G, Keul J. The significance of the aerobicanaerobic transition for the determination of work load intensities during endurance training. Eur J Appl Physiol 1979; 42: 25 34 13 Metra M, Raddino R, Dei Cas L, Visioli O. Assessment of peak oxygen consumption, lactate and ventilatory thresholds and correlation with resting and exercise hemodynamic data in chronic congestive heart failure. Am J Cardiol 1990; 65: 1127 1133 14 Montgomery DL, Baudin PA. Blood lactate and heart rate response of young female during gymnastic routines. J Sports Med 1982; 22: 358 365 15 Spriet LL. Anaerobic metabolism during high-intensity exercise. In: Hargreaves M (ed). Exercise Metabolism. Champaign, IL: Human Kinetics, 1995: 13 Corresponding Author: Laura Guidetti I.U.S.M. / Istituto Universitario di Scienze Motorie Piazza Lauro De Bosis, 15 00194 Roma Italy Phone: + 39 (6) 36095599 Fax: + 39 (6) 3613065 E-mail: la.guidetti@confor.it