Science of Gymnastics Journal vol. 11, num. 2, year 2019 Published by Department of Gymnastics, Faculty of Sport, University of Ljubljana ISSN 1855-7171 Science of Gymnastics Journal (ScGYM®) Science of Gymnastics Journal (ScGYM®) (abrevated for citation is SCI GYMNASTICS J) is an international journal that provide a wide range of scientific information specific to gymnastics. The journal is publishing both empirical and theoretical contributions related to gymnastics from the natural, social and human sciences. It is aimed at enhancing gymnastics knowledge (theoretical and practical) based on research and scientific methodology. We welcome articles concerned with performance analysis, judges' analysis, biomechanical analysis of gymnastics elements, medical analysis in gymnastics, pedagogical analysis related to gymnastics, biographies of important gymnastics personalities and other historical analysis, social aspects of gymnastics, motor learning and motor control in gymnastics, methodology of learning gymnastics elements, etc. Manuscripts based on quality research and comprehensive research reviews will also be considered for publication. The journal welcomes papers from all types of research paradigms. Editor-in-Chief Ivan Čuk, Slovenia Responsible Editor Maja Pajek, Slovenia Science of Gymnastics Journal is indexed in Web of Science (ESCI data base, since 2015), EBSCOhost SPORTDiscus, SCOPUS, COBISS Editorial and Scientific Board Koichi Endo, Japan Marco Antonio Bortoleto, Brazil Nikolaj Georgievic Suchilin, Russia William Sands, USA Kamenka Zivcic, Croatia Ignacio Grande Rodríguez, Spain Warwick Forbes, Australia David McMinn, Scotland, UK Almir Atikovic, Bosnia and Herzegovina José Ferreirinha, Portugal Istvan Karacsony, Hungary Hardy Fink, FIG Academy, Canada Keith Russell, FIG Scientific Commission, Canada Thomas Heinen, Germany Front page design: Sandi Radovan, Slovenia. Editorial Office Address Science of Gymnastics Journal Faculty of Sport, Department of Gymnastics Gortanova 22, SI-1000 Ljubljana, Slovenia Telephone: +386 (0)1 520 7765 Fax: +386 (0)1 520 7750 E-mail: scgym@fsp.uni-lj.si Home page: http://www.scienceofgymnastics.com (IZUM), SIRC (Canada), ERIHPLUS, OPEN. J-GATE, GET CITED, ELECTRONIC JOURNALS SCHOLAR, PRO QUEST and INDEX COPERNICUS. ScGYM® (ISSN 1855-7171) is an international online journal published three times a year (February, June, October). ® Department of Gymnastics, Faculty of Sport, University of Ljubljana. All rights reserved. This journal and the individual contributions contained in it are protected under Copyright and Related Rights Act of the Republic of Slovenia. INDEX, SCIRUS, NEW JOUR, GOOGLE Science of Gymnastics Journal is supported by Foundation for financing sport organisations in Slovenia, Slovenian Research Agency and International Gymnastics Federation. Slovenian Research Agency SCIENCE OF GYMNASTICS JOURNAL Vol. 11 Issue 2: 2019 CONTENTS Ivan Cuk EDITORIAL 137 Sarita Bacciotti Adam D.G. Baxter-Jones Adroaldo Gaya José Maia MOTOR PERFORMANCE OF BRAZILIAN FEMALE ARTISTIC GYMNASTS: INSIGHTS VIA MULTILEVEL ANALYSIS 139 Dallas George Alkmini Savvathi Konstantinos Dallas Maria Maridaki THE EFFECT OF 6-WEEKS WHOLE BODY VIBRATION ON MUSCULAR PERFORMANCE ON YOUNG NON-COMPETITIVE FEMALE ARTISTIC GYMNASTS 151 Ole H. Hansen Lars G. Hvid Per Aagaard Kurt Jensen MECHANICAL LOWER LIMB MUSCLE FUNCTION AND ITS ASSOCIATION WITH PERFORMANCE IN ELITE TEAM GYMNASTS 163 Iliya Kiuchukov Iliya Yanev Lubomir Petrov Stefan Kolimechkov Albena Alexandrova Dilyana Zaykova Emil Stoimenov IMPACT OF GYMNASTICS TRAINING ON THE HEALTH-RELATED PHYSICAL FITNESS OF YOUNG FEMALE AND MALE ARTISTIC GYMNASTS 175 Toshiyuki Fujihara Pierre Gervais Gareth Irwin HEAD-TOE DISTANCE AS A SIMPLE MEASURE TO EVALUATE AMPLITUDE OF CIRCLES ON POMMEL HORSE 189 Lucija Milcic Kamenka Zivcic Tomislav Krsticevic DIFFERENCES IN VAULT RUN-UP VELOCITY IN ELITE GYMNASTS 201 Jonas Rohleder Tobias Vogt EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND PERFORMANCES IN NOVICES: INVERTING EXPLICIT AND IMPLICIT LEARNING OF SKILL-RELATED MOTOR TASKS 209 William A. Sands Madison K. Varmette Gregory C. Bogdanis Olyvia Donti Bryce V. Murphy Troy J. Taylor COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES OF TRAMPOLINES 223 Miriam Kalichovâ Petr Hedbâvny Barbora Pyrochtovâ Jana Prihonskâ COMPARISON OF ACTUAL AND PREDICTED ANTHROPOMETRIC CHARACTERISTICS OF CZECH ELITE GYMNASTS 239 Adriana Kaplânovâ PERSONALITY OF GYMNASTS AND COPING STRATEGIES TO MANAGE STRESS 255 Anton Gajdos Michal Bâbela HISTORICAL SHORT NOTES XV 266 SLOVENSKI IZVLEČKI / SLOVENE ABSTRACTS 269 135 SCIENCE OF GYMNASTICS JOURNAL Vol. 11 Issue 2: 2019 13th Freiburg International Gymnastics Congress 19-20 October 2019 Albert-Ludwigs-Uni vers i läi Fre ihurg 20 Top Speakers 60 Talks, Lectures and Workshops License renewal for all license levels 135 SCIENCE OF GYMNASTICS JOURNAL Vol. 11 Issue 2: 2019 EDITORIAL Dear friends, The European Union has changed its approach toward scientific publications and we have to respect the new guidelines. The most important guideline is that all scientific articles have to be openly accessible. While our journal already has open access to articles, we will try to number articles by digital object identifier (DOI) by the end of the year. It will be slightly more work, but generally we will still be able to publish three issues per year. At the end of May new journal evaluations have been published in SCOPUS. Unfortunately, our citation has been placed slightly lower than last year, but our SNIP indicator has risen and thus our journal is now in the second quarter of journals. An excellent result! In this issue, we have again ten articles by authors from Brazil, Portugal, Canada, Greece, Denmark (for the first time), Bulgaria (for the first time), Great Britain, Japan, Croatia, Germany, the USA, the Czech Republic and Slovakia. There is a variety of research fields and it is good to see that there is a lot of productive international cooperation among researchers. Among articles on gymnastics disciplines, most are dealing with the man and the women artistic gymnastics; we are proud that for the first time we have an article from TEAMGYM, a discipline that is gaining momentum in Europe. Anton Gajdos prepared another article related to the history of gymnastics, refreshing our information on Albert Azarjan, an excellent Armenian (ex Soviet Union) gymnast. Please be welcome to Freiburg to 13th International Gymnastics Congress. Just to remind you, if you quote the Journal, its abbreviation on the Web of Knowledge is SCI GYMN J. I wish you pleasant reading and a lot of inspiration for new research projects and articles, Ivan Cuk Editor-in-Chief 135 SCIENCE OF GYMNASTICS JOURNAL Vol. 11 Issue 2: 2019 135 Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 MOTOR PERFORMANCE OF BRAZILIAN FEMALE ARTISTIC GYMNASTS: INSIGHTS VIA MULTILEVEL ANALYSIS Sarita Bacciotti 12', Adam D.G., Baxter-Jones4, Adroaldo Gaya5, José Maia1 1 Faculty of Sport, University of Porto, Portugal 2 Physical Education Course, Federal University of Mato Grosso do Sul, Campo Grande, Brazil 3CAPES Foundation, Ministry of Education of Brazil, Brasilia—DF, Brazil 4 College of Kinesiology, University of Saskatchewan, Canada 5 Department of Physical Education, Federal University of Rio Grande do Sul, Porto Alegre, Brazil. _Original article Abstract The aim of this research was to investigate individual and club-level variables that explain individual differences in gymnasts' motor performance (MP). The sample was comprised of 249 female gymnasts (68 elite; 181 non-elite), aged between 9 and 20 years, split into four age categories: 9-10 years (n=98); 11-12 years (n=72); 13-15 years (n=64), and 16 and above (n=15). Gymnasts were from 26 Brazilian clubs, from six different states. The Talent Opportunity Program physical ability total test score was used to assess gymnasts' MP, based on a battery of seven tests: handstand hold, cast, rope climb, press handstand, leg flexibility, leg lift, and 20 meter sprint. Anthropometric, body composition, biological maturation, and training history data were also collected, as were club dimensions, infrastructures, competitions, manpower, and availability of selection/talent programs. Data were analyzed using a multilevel modelling approach. Individual gymnast' characteristics explained 39% of physical ability score variance from which 32% was related to the independent effects of age, competitive level, fat free mass, occurrence of menarche, and trainings hours per week (p<0.05). Club characteristics explained 61% of gymnasts' total variance in physical ability score; 96% of this amount was related to club dimension, manpower, and talent program. These results reinforce the relevant role of the contextual effects and highlight the need to invest in club infrastructures: ideally in coaches' expertise and effective selection programs. Such investments should enable the enhancement and development of a gymnast's careers during their lifetime involvement in training and competition. Keywords: physical fitness, performance, gym club. INTRODUCTION Elite women artistic gymnasts (WAG) differ from their peers in their: physique aesthetics, explosive power and specialized skills, and their technical perfection in each of the elements (Bradshaw, Hume, & Aisbett, 2011). Although gymnast's high performance levels may express different genetic endowments (Morucci, Punzi, Innocenti, Gulisano, Ceroti, & Pacini, 2014), it is also acknowledged that performance is related to training conditions, continuous Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 competition engagement, and coaching excellence (Côté, 1999). It has always been challenging to assess gymnasts' motor performance (MP), apart from specialized skill-oriented tasks (Marina, Jemni, & Rodriguèz, 2013). A gymnast's physical attributes (Bajin, 1987; Vandorpe, Vandendriessche, Vaeyens, Pion, Lefevre, Philippaerts et al., 2011) interact with, and reflect, motor components (Sands, 2011), with the greatest emphasis's placed on strength and explosive power, flexibility and artistry (Bale & Goodway, 1990). The putative MP measurements are made difficult by having a complex set of parameters such as technical skills, muscular contractions and speed of stretch (Monèm, 2011). However, test batteries to assess MP have been developed to identify specific gymnastics motor profiles (Albuquerque & Farinatti, 2007), and to identify giftedness amongst gymnasts (Bajin, 1987; USA-Gymnastics, 2014). The new 2014 Talent Opportunity Program (TOP) offers a comprehensive test battery for gymnastics and it has been suggested that the resulting scores (TOPS total score - TTS) could be used in talent identification programs (USA-Gymnastics, 2014). Since gymnasts spend considerable amounts of their sport life time training in their clubs, it is likely club infrastructures, competition schedules, number and quality of coaches, as well as selection programs may affect their MP. For example, manager/coach, pay/salary and training conditions ((Kayani, Zia, & Abbas, 2012), human resources in achieving organizational objectives, the relationship between human resources and sports performance ((Mihaela, Veronica, & Dana, 2014), and the joint effects of individual and group level covariates (Hill, Stoeber, Brown, & Appleton, 2014; (Petitta, Jiang, & Palange, 2015) have all been shown to be related to a gymnasts performance. The aim of this research was to investigate individual and club-level variables that explain individual differences in gymnasts' motor performance (MP). METHODS The sample comprises 249 female gymnasts [68 elite (EG); 181 non-elite (NEG)] aged 9-20 years. They were separated into four age categories according to the Brazilian Gymnastics Federation competition rules (CBG, 2015): 9-10 years (n=98); 11-12 years (n=72); 1315 years (n=64), and >16 yeas (n=15). Gymnast were classified as EG or NEG using the following criteria: NEG are those who participated in regional or state championships, or competed at the national championship but were classified below the 10th position (all-around classification); EG are those who either participated in national championship and had been classified between the 1st and the 10th position in the final ranking (all-around classification), or who had participated in international championships. Gymnasts were selected from 26 Brazilian gymnastics clubs, representing ~60% of all clubs in a state. Clubs were selected based on their participation and classification in the 2014 Official Brazilian Championships. All gymnasts included in the study (Table 1) were identified by their coaches and were part of the main team in each club. The study protocol was approved by the ethics committee of Dom Bosco Catholic University (CAAE 42967215.9.0000.5162), as well as by the technical director of all the Gymnastics Clubs. Written consent forms were obtained from parents or legal guardians of gymnasts, as well as assent from all gymnasts. The Talent Opportunity Program physical ability testing score (TTS), (USA-Gymnastics, 2014) was calculated from a battery of seven MP test: handstand hold; cast; rope climb; press handstand; leg flexibility; leg lift; and 20 meter sprint. Each of the 7 test scores was scored Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 between 0-10 points, except leg flexibility which was scored between 0-12 points. The upper limit of the TTS is 72 points, and a higher score indicates better MP. All anthropometric measurements were made according to standardized protocols (Ross, & Marfell-Jones, 1991). Height and sitting height were measured to the nearest 0.1cm using a portable stadiometer (Sanny Stadiometer, SP, BR) with the head positioned in the Frankfurt plane. Body mass (Kg) was measured with a portable bio-impedance scale (Tanita SC 240 Body Composition Analyser scale, IL, USA) with a 0.1kg precision. Leg length was calculated as the difference between standing height and sitting height. All measurements were performed at the beginning of each training session. Body composition was estimated from regression equations provided by the manufacturer of the bio-impedance Tanita SC 240 scale (Tanita SC 240 Body Composition Analyser scale, IL, USA) which were unavailable to researchers. In the present study, fat mass (Kg), free fat mass (Kg), and percent fat (%Fat) were considered. Biological maturity was obtained using data from menarche occurrence (yes or no) as well as from predicting age from (maturity offset) the attainment of peak height velocity (PHV) using a anthropometric variables (Mirwald, Baxter-Jones, Bailey, & Beunen, 2002). In girls, the predictive equation is: Maturity Offset=-9.376 + 0.0001882 * leg length and sitting height interaction + 0.0022 * age and leg length interaction + 0.005841 * age and sitting height interaction -0.002658 * age and weight interaction + 0.07693 * weight by height ratio. A positive (+) maturity offset indicates the number of years the participant is beyond PHV, whereas a negative (-) maturity offset represents the number of years the participant is prior to the attainment of PHV. Athletes answered questions regarding their onset of training, training years, onset of competition, as well as the number of training hours per week. These answers were cross-checked with their parental as well as coaches' reports. When inaccuracies were found, we relied on the information provided by the coach. Information about all gymnastic clubs was obtained via questionnaire which was developed in accordance with an expert panel of gymnastics coaches and researchers from Brazilian and Portuguese clubs and Sport Sciences colleges. It was centered on 5 domains: (1) club dimensions; (2) infrastructures; (3) competition; (4) manpower; (5) selection/talent programs. Club managers or persons with similar functions answered all questions. Club dimensions (CD) recorded information about the total number of women artistic gymnasts (WAG) practicing at the club. A dummy coding schema was used to categorize the responses: CD=0 if up to 50 gymnasts (reference category); CD=1 if between 51 to 150 athletes; CD=2, if 150+ gymnasts. Further, the number of gymnasts per team was assessed (GPT). GPT=0 if gymnasts <6 (reference category); GPT=1 if 7 to 10 gymnasts; and GPT=2, if >10 gymnasts. Infrastructures included questions about the exclusive use of the club for gymnastics, availability of complete sets of WAG apparatuses and Pit (foams' pool for security used in the end of or under gym apparatuses) in the gym. These variables were binary coded 1=yes or 0= no. In this domain questions were related to WAG lifetime (LT) in the club. This was coded in the following way: LT=0 if up to 4 years (reference category), LT=1 if 5 to 10 years, and LT=2 if >11 years. Participation in international competitions was also recorded, binary coded as 1=yes or 0=no. Manpower included questions about the amount of WAG coaches (C): C=0 if <=to 4 coaches, and C=1 if >4 coaches. Further, coaches experience (CE) was also recorded in years, in competitive WAG: Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 CE= 0 if <=4 years, CE=1 if >4 but <10 years, CE=2 if >10 years but < 20 and CE = 3 if >20 years. In this domain questions were asked about the way the gymnasts' search/selection was made, because most selections are made external or internal to the club. As such our reference variable (0) was if selection (S) made externally in groups of children interested to participate in gymnastics: S=1 if external from club but from children already participating in gymnastics; S=2 if internal to the club, and S=3 if different from the previous. A second question was about the systematic use of tests in selection, coded as 0 (reference category, meaning "no") and 1 (yes, they use). Exploratory and descriptive statistics were performed in SPSS 20.0. Mean differences between EG versus NEG, in each age group, were calculated in STATA 14, using a t-test with a Satterthwaite's approximation for degrees of freedom, since in most cases the variances were not equal between the groups. Alpha was set at p=0.01. Mean differences for relative (%) data were performed in WinPep software (Abramson, 2004). Since our data was clustered, i.e., gymnasts (level-1) nested within clubs (level-2), we used a hierarchical multilevel model and data were analyzed using Supermix v.1 software (Hedeker, D, Gibbons, R, du Toit, M, & Cheng, Y, 2008). A sequence of nested models was developed and tested as is common in multilevel modelling (Hox, 2010). We started with a simple model and then added predictors, i.e., we increased the model complexity. Simpler models are contained in more complex models, i.e., they are nested within. Deviance is the usual statistic that describes the goodness of fit of a model. More complex models in terms of explanatory power as well as in number of parameters, i.e., with more predictors, are expected to have lower values of Deviance. Differences in Deviances follow a Chi-Square distribution with degrees of freedom equal to the difference in the number of parameters of both models (Hox, 2010). Further, the relevancy of individual and club-level predictors to explain TTS variance was assessed with a pseudo-R2 (which is similar, in a way, to the R2 of multiple linear regression), and is defined as the proportional reduction in variance resulting from a comparison of a new model with a previous one (Hedeker et al., 2008). Modeling was performed in a sequential fashion as is usual when using any statistical model to explain any outcome variable - in our case TTS. First a Null model (M0) was estimated to compute the intraclass correlation, i.e., the variance accounted for by clubs' effects on gymnasts' TTS - the main issue here is to answer the fundamental question: are clubs important in explaining gymnasts' TTS scores? Then, in Model 1 (M1) we only included gymnasts' TTS predictors. Finally, in Model 2 (M2), we added club predictors. This is, in fact, our most complex model in terms of explanatory power and also in the number of parameters to be estimated. RESULTS Gymnasts' descriptive statistics (means ± standard deviations and percentages) are shown in Table 1. As expected, the EG had significantly higher TTS's (p<0.01). However, there were no differences in size, maturation or body composition between gymnasts at any age group (p>0.01) apart from significantly more NEG 13-15 year old's attaining menarche (p<0.01). There were also no differences in the age of training onset or training years between the groups (p>0.01), however the GE, 11-12 and 1315 year old's performed significantly more training hours per week (p<0.01). Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF. Vol. 11 Issue 2: 139 - 150 Table 1 Descriptive statistics by level of gymnastics (non-elite -NEG and elite-EG) and age group. 9-10 yrs 11-12 yrs 13-15yrs > 15 yrs n=98 n=72 n=64 n=15 Competitive Level NEG EG NEG EG NEG EG NEG GE n=84 n=14 n=45 n=27 n=41 n=23 n=11 n=4 Age (yrs) 9.49±0.50 9.86±0.36 11.51±0.51 11.52±0.51 13.76±0.73 13.57±0.66 17.00±0.78 17.50±1.92 TTS 26.68±13.11* 45.07±13.60 25.47 ±13.90* 48.11±13.99 34.96±15.54* 53.52±10.10 30,63±13,44 52,66±7,68 Weight (kg) 29.03±3.88 30.84±4.19 37.59±7.37 35.20±6.97 47.37±7.65 43.54±4.95 53.78±5.11 52.40±4.06 Height (cm) 134.69±6.22 135.76±5.91 146.12±7.16 142.64±6.60 154.23±6.98 150.84±4.76 159.33±4.28 159.18±4.97 Free Ft Mass (kg) 24.17±2.92 25.30±3.03 30.37±4.34 28.66±3.98 36.53±4.60 34.37±2.99 42.72±2.54 40.96±1.75 Maturity Offset -2.71±0.49 -2.37±0.47 -1.07±0.60 -1.28±0.60 0.61±0.79 0.29±0.57 2.76±0.50 3.00±1.27 Menarche occurrence 3 (3.60%) - 5 (11.10%) 3 (11.10%) 30 (73.20%)* 7 (30.40%) 11(100%) 4(100%) Training onset (yrs) 5.84±1.53 5.07±1.33 6.55±2.05 5.82±1.27 6.40±2.01 6.00±1.41 8.09±2.38 5.50±1.91 Training years (yrs) 3.89±1.91 5.11±1.60 5.13±2.19 5.92±1.75 7.43±2.32 7.91±1.56 9.27±2.80 11.75±4.71 Training hours (h-w-1) 20.66±7.03 24.39±4.35 19.40±8.88* 28.20±3.09 21.7±8.8* 30.13±4.21 24.00±8.31 27.75±3.50 Competition onset (yrs) 7.52±1.51 6.85±1.19 8.05±1.69 7.59±1.21 8.24±1.54 7.61±1.23 9.27±1.90 8.25±0.50 *p<0.01 Science of Gymnastics Journal 143 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 Table 2 Variable domains and their results at the club level. Club dimension Total Number of Gymnasts - WAG(1) Up to 50 gymnasts 51 - 150 gymnasts Over 150 gymnasts Number of Gymnasts per team Less than 6 gymnasts 7 - 10 gymnasts More than 11 gymnasts_ Infrastructures Exclusive place for gymnastics No Yes Complete Apparatuses of WAG and Pit(2) in the gym No Yes_ Competition Lifetime of WAG in the club Up to 4 years 5 to 10 years Over 11 years Gymnasts with participation in International Competitions No Yes_ Manpower Amount of WAG coaches in teams Up to 4 coaches 5 to 10 coaches Coach time experience in competitive WAG Up to 4 years 5 to 10 years 11 to 20 years More than 20 years_ Talent programs The way in which the selection of gymnasts is performed. Externally (evaluating girls still no exercise practitioners) 7.70% Externally (evaluating girls who already practice or practiced gymnastics) 15.40% Internal form (evaluating girls already practicing gymnastics at the club) 53.80% Another (not identified way). 23.10% Use of tests in gymnasts selection No 15.40% Yes 84.60% WAG(1) = Women's Artistic Gymnastics, Pit(2)=security place for training 23.00% 38.50% 38.50% 34.60% 42.30% 23.10% 7.70% 92.30% 30.80% 69.20% 11.50% 15.40% 73.10% 50.00% 50.00% 92.30% 7.70% 15.40% 34.60% 26.90% 23.10% Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF. Vol. 11 Issue 2: 139 - 150 Table 3 Results summary of hierarchical linear modeling. Parameters Null Model Model 1 Model 2 Estimates SE p Estimates SE(1) p Estimates SE p Fixed Effects Gymnasts Level Intercept 34.80 2.72 <0.01 27.70 1.85 <0.01 42.02 6.77 <0.01 Age Categories 4.17 1.31 <0.01 4.07 1.33 <0.01 Competitive Level 11.16 2.28 <0.01 11.98 2.25 <0.01 Free Fat Mass (kg) -0.34 0.13 <0.01 -0.36 0.13 <0.01 Menarche occurrence 11.57 4.43 <0.01 11.54 4.38 <0.01 Training hours/per week (h-w-1) 1.12 0.13 <0.01 1.22 0.13 <0.01 Age Cat(2)-by-Comp Level(3) Interaction -2.21 1.59 >0.01 -3.01 1.58 >0.01 Menarche occurrence-by-Age Cat Interaction -4.46 2.13 >0.01 -4.47 2.12 >0.01 Club Level Club dimension Total WAG: 51 -150 -7.15 3.31 >0.01 Total WAG: over 150 3.94 2.75 >0.01 Gymnasts per team: 7 - 10 gymnasts -4.16 2.21 >0.01 Gymnasts per team: More than 11 gymnasts -9.32 2.60 <0.01 Manpower Number of WAG coaches in teams -7.99 4.06 >0.01 Coach time experience 5 to 10 yrs 0.31 3.38 >0.01 11 to 20 yrs3 7.40 3.61 >0.01 More than 20 yrs 4.25 3.11 >0.01 Talent/Selection programs Selection externally (already practitioners) -21.55 6.22 <0.01 Selection internally at the club -12.11 4.73 <0.01 Another selection process (not identified) -12.12 4.83 <0.01 Use of tests in gymnasts selection 7.41 2.84 <0.01 Random effects Intercept 177.32 39.43 7.20 Residual 115.95 79.19 79.38 Deviance (number of parameters) 1951.44 (3) 1802.53 (10) 1775.28 (22) SE(1)= Standard Error, Age Cat(2)= Age Categories, Comp Level(3)= Competitive Level, WAG(4)=Women's Artistic Gymnastics. Science of Gymnastics Journal 143 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 Clubs' descriptive statistics are shown in Table 2. Nearly 80% of clubs had more than 50 gymnasts and at least 4 coaches with between 5 and 20 years of experience. A third of the clubs had 6 gymnasts in a team and over a half had athletes participating in international competitions, with the majority having competed for over 11 years. Most clubs were exclusively for gymnastics and over two thirds had a pit and complete sets of apparatuses. The most frequent mode of selection into the sport was internal to the club (54%), and most clubs (85%) used tests to identify talent. The results of the multilevel models are shown in Table 3. In the Null Model (M0), the intercept (fixed effects) represents the average TTS points (P=34.8±2.7) for all gymnasts. Variance between gymnasts and between clubs is shown in the random effects; the intraclass correlation coefficient (ICC)=0.61, implies that 61% of the total variance in gymnasts' TTS is explained by club level covariates; further, the remaining variance, 39%, is explained by differences in individual gymnasts' characteristics. M1 fitted the data better than M0 indicated by a reduction in goodness of fit from 1951.4 to 1802.5 (p<0.01). The intercept of M1 (p=27.7±1.9, p<0.01) represents the TTS for a gymnast from the reference group (9-10 yrs old). TTS increased with increasing age category (P=4.2±1.3, p<0.01); further, EG gymnasts scored higher on TTS (p=11.2±2.8, p<0.01) when all other confounders were controlled. Gymnasts with higher free fat mass performed worst in TTS (P=-0.3±0.1, p<0.01). The higher the number of training hours per week the better the performance (P=1.1±0.1, p<0.01). Experiencing menarche was positively associated with TTS scores (p=11.6±4.4, p<0.01). Taken together, these gymnasts' covariates explained 32% of the 40% inter-individual TTS scores variance. Since most clubs have similar infrastructures, as well as competition no association was found with TTS (p>0.01). M2, with club-level covariates added fitted the data better than M1. In M2, the intercept (p=41.6±6.85, p<0.01) is the TTS of a gymnast when all predictors are at zero. Gymnast level covariates remained similar to those in M1. Results from M2 showed that gymnasts from small clubs (less than 50 gymnasts) and small number of gymnasts in team classes, as well as from clubs with less than 4 coaches but with more experience time in training tend to perform better in TTS (p<0.01). Further, gymnasts from clubs that had selection tests and who selected gymnasts externally (no practitioners) tend to perform better in TTS (p<0.01). Taken together, club-level covariates explained 96% of the ~61% of the between-clubs variance. DISCUSSION This paper aimed to identify gymnast and club level characteristics that explained inter-individual differences in motor performance (MP) tests. The main findings indicated that a substantial amount of TTS variance (61%) was attributable to club environment and to a smaller extent (39%) the individual gymnast's characteristics. Total TTSs score increases of ~4.2 points were observed across age categories, this is similar to a mean increase of ~3 points across 9 to 11 years old previously reported in USA gymnasts (USA-Gymnastics, 2010). Further, a previous study of Brazilian gymnasts, albeit using a different test battery (Albuquerque, & Farinatti, 2007), showed mean increments of 2.5 points among beginners and 3.7 points among elite gymnasts aged 9 to 15 years old. These results confirm the roles of both neuromuscular maturation and experience in performing strength tasks (Malina, Bar-Or, & Bouchard, 2004). In 2010, the American Gymnastics Federation reported TTS averages of 53.47, 56.43 and 59.57 points in 9, 10 and 11 yrs old gymnasts Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 respectively, which are greater than those reported in present sample of Brazilian gymnasts. This may explain in part their higher performances. In our models a constant difference in EG and NEG TTS of ~11.2 points was evident across all age categories, favoring the EG. A similar trend was previously reported between beginners and elite Brazilian gymnasts, but again using a different test battery (Albuquerque, & Farinatti, 2007). (Vandorpe et al., 2011) also found significant differences in physical performance characteristics between nonelite and elite levels, favoring the elite group, even after controlling for age and maturation (age at peak height velocity). These consistent differences could be explained in part by the greater amount of weekly training hours performed by EG. There was a positive effect of biological maturation on TTS of ~11.6 points, i.e., a superior MP of those who already passed through menarche. However, a statistically significant interaction between menarcheal status and age categories revealed a decrease of ~4.5 TTS points. It has consistently been shown that EG are late maturers (Baxter-Jones, Thompson, & Malina, 2002; Malina, Baxter-Jones, Armstrong, Beunen, Caine, Daly et al., 2013), and since a high frequency of NEG reached menarche prior to EG in each age category this may help explain this interaction. As expected, the number of training hours per week was positively associated with TTS performance, on average, ~1.2 points. This is in agreement with available data (Vandorpe et al., 2011), given that weekly time in training increases with age and competition level (Malina et al., 2013). As per (Ericsson, Krampe, & Tesch-Romer, 1993) suggestion, more training hours per week provide more deliberate practice which then translates to higher levels of expertise, which is consistent with (Lidor, Tenenbaum, Ziv, & Issurin, 2016) findings. Results from club-level covariates showed significant associations of club dimensions, manpower and talent programs with gymnasts' TTS development. Gymnasts from larger clubs (over 150 gymnasts) had better performance while those from medium clubs (51 - 150 gymnasts) had worse performance (-7.0) in the TTS. It is possible that these results may be due to the fact that larger clubs have a larger pool to select from and have better opportunities to train them. In contrast, smaller clubs work with those available, who may be less gifted. Team size was negatively associated with performance suggesting smaller teams are prone to more individualized planning, coaches' attention and training together with more systematical assessments and skills enhancement (Asqalan, 2016). In educational settings classes with more students are thought to have a negative influence in educational attainment (Case, & Deaton, 1999), whist small classes, especially in the early grades, lead to higher academic achievement (Nye, Hedges, & Konstantopoulos, 2000). Translating this to gymnastics training, where quality and repetition are very important, more individual attention allows the gymnast to develop a high degree of perfection (Asqalan, 2016). Gymnasts belonging to clubs with more coaches had worse TTS performance (-8.4 points), at the same time more coaching experience was positively associated with TTS. (Baker, Horton, Robertson-Wilson, & Wall, 2003) highlighted that the coach is most probably one of the utmost significant keys to athlete development and performance. Additionally, more years of experience allow coaches to draw upon their vast and diverse amounts of information about their sport as well as their athletes, since more experience allows better planning, diagnose, and strategize more effectively (De Marco, & Mccullick, 1997). Since the development of expertise is a long-term Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 process, coaches who achieved it are more efficient in detecting what athletes need to know and then find the best strategies to supply that information to them (De Marco, & Mccullick, 1997). As is to be expected, clubs with selection processes had higher TTS scores (Albuquerque, & Farinatti, 2007; Bajin, 1987; Pion, Lenoir, Vandorpe, & Segers, 2015) as did those who teams were selected externally. This is likely associated with a broader recruitment basis together with the fact that decisions are mostly based on innate characteristics that are believed to be mandatory to excel, and not so much on specific technical skills (Meyers, Van Woerkom, & Dries, 2013). This study has limitations, specifically the sample does not represent all Brazilian gymnasts and care needs to be taken about the generalizability of the findings. However, the sample includes gymnasts from states with higher competitive levels across Brazil. Secondly, the TOPS battery is not widely implemented in Brazil and its importance and applicability in gymnasts' training control is unknown. CONCLUSIONS The present study showed that individual gymnast' traits explained 40% of TTS variance, from which 32% was related to the independent effects of age, competitive level, fat free mass, menarche occurrence, and trainings hours per week, i.e., older, lighter and mature gymnasts, as well as those belong to the elite group that train more hours per week perform better. Moreover, club characteristics explained 61% of gymnasts' total variance in TTS performance, and 96% of this amount is related club dimension, manpower, and talent program, reinforcing the relevant roles of contextual effects, which highlights the need to also invest in club structures, mainly in coaches' expertise as well as in effective selection programs to develop and enhance gymnasts' carriers during their lifetime involvement in serious training and competition. ACKNOWLEDGMENT This work was supported by CAPES Foundation, Ministry of Education of Brazil, Brasilia—DF, Brazil, (11927/13-5/PhD scholarship). REFERENCES Abramson, Joseph H. (2004). WINPEPI (PEPI-for-Windows): computer programs for epidemiologists. Epidemiologic Perspectives & Innovations, 1(1), 6. Albuquerque, Patricia Albuquerque, & Farinatti, Paulo Tarso Veras. (2007). Development and validation of a new system for talent selection in female artistic gymnastics: The PDGO Battery. Revista Brasileira de Medicina do Esporte, 13(3), 157-164. Asqalan, Manar. (2016). The effect of EFL large classes on Yarmouk University Students' achievement. [Article]. Arab World English Journal, 7(1), 358-369. Bajin, Boris. (1987). Talent identification program for Canadian female gymnasts. In Petiot, Bernard, Salmela, John H & Hoshizaki, T Blaine (Eds.), World Identification Systems for Gymnastic Talent (pp. 34-44). Montreal: Sport Psyque Editions. Baker, Joseph, Horton, Sean, Robertson-Wilson, Jennifer, & Wall, Michael. (2003). Nurturing sport expertise: factors influencing the development of elite athlete. J Sports Sci Med, 2(1), 1-9. Bale, Peter, & Goodway, Jackie. (1990). Performance variables associated with the competitive gymnast. / Variables de performance du gymnaste de competition. Sports Medicine, 10(3), 139145. Baxter-Jones, Adam D G, Thompson, Angela M, & Malina, Robert M. (2002). Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 Growth and maturation in elite young female athletes. Sports Medicine and Arthroscopy Review, 10(1), 42-49. Bradshaw, Elizabeth Jane, Hume, Patria Anne, & Aisbett, Brad. (2011). Performance score variation between days at Australian national and Olympic women's artistic gymnastics competition. Journal of Sports Sciences, 30(2), 191-199. Case, Anne, & Deaton, Angus. (1999). School inputs and educational outcomes in South Africa. [Article]. Quarterly Journal of Economics, 114(3), 1047-1084. CBG. (2015). Capitulo V - Das Categorias [Chapter V - Categories] Regulamento Geral Confederaçâo Brasileira de Ginâstica [Categories General Regulation of Gymnastics Brazilian Federation] (pp. 6-7). Aracaju-BR: Confederaçâo Brasileira de Ginâstica. Côté, Jean. (1999). The influence of the family in the development of talent in sport. [Article]. Sport Psychologist, 13(4), 395-417. De Marco Jr, Georg M, & Mccullick, Bryan A. (1997). Developing expertise in coaching: Learning from the legends. Journal of Physical Education, Recreation & Dance, 68(3), 37-41. Ericsson, K Anders, Krampe, Ralf T, & Tesch-Romer, Clemens. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363. Hedeker, Donald, Gibbons, Robert, du Toit, Mathilda, & Cheng, Yan. (2008). Manual of SuperMix: mixed effects models. Lincolnwood, IL: Scientific Software International. Hox, Joop J. (2010). Multilevel Analysis. Techniques and Applications (2nd ed.). New York: Routledge. Kayani, Jawad, Zia, M Qamar, & Abbas, Thair. (2012). What variables shape player's performance in professional teams/clubs. Journal of Physical Education & Sport, 12(3), 380-384. Lidor, Ronnie, Tenenbaum, Gershon , Ziv, Gal, & Issurin, Vladimir. (2016). Achieving expertise in sport: deliberate practice, adaptation, and periodization of training. Kinesiology Review, 5(2), 129141. Malina, Robert M, Bar-Or, Oded, & Bouchard, Claude. (2004). Growth, maturation, and physical activity (2nd ed.). Champaign: Human Kinetics. Malina, Robert M, Baxter-Jones, Adam D G, Armstrong, Neil, Beunen, Gaston P, Caine, Dennis, Daly, Robin M, et al. (2013). Role of intensive training in the growth and maturation of artistic gymnasts. Sports Medicine, 43(9), 783802. Marina, Michel, Jemni, Monèm, & Rodriguèz, Ferran A. (2013). Jumping performance profile of male and female gymnasts. Journal of Sports Medicine and Physical Fitness, 53(4), 378-386. Meyers, M Christina, Van Woerkom, Marianne, & Dries, Nicky. (2013). Talent — Innate or acquired? Theoretical considerations and their implications for talent management. [Article]. Human Resource Management Review, 23, 305321. Mihaela, Iconomescu Teodora, Veronica, Mindrescu, & Dana, Badau. (2014). Importance of human resources in sports clubs and in sports performance management. Gymnasium: Scientific Journal of Education, Sports & Health, 15(2), 269-278. Mirwald, Robert L, Baxter-Jones, Adam D G, Bailey, Donald A, & Beunen, Gaston P. (2002). An assessment of maturity from anthropometric measurements. [Article]. Medicine and Science in Sports and Exercise, 34(4), 689694. Monèm, Jemni. (2011). Specific physical and physiological assessments of gymnasts. In Jemni, Monèm (Ed.), The science of gymnastics (pp. 32-38). London: Routledge. Morucci, Gabriele, Punzi, Tiziana, Innocenti, Giovanni, Gulisano, Massimo, Ceroti, Marco, & Pacini, Stefania. (2014). New frontiers in sport training: genetics Science of Gymnastics Journal 139 Science of Gymnastics Journal Baciotti S., Baxter-Jones A.D.G., Gaya A., Maia J.: MOTOR PERFORMANCE OF... Vol. 11 Issue 2: 139 - 150 and artistic gymnastics. Journal of Strength and Conditioning Research, 28(2), 459-466. Nye, Barbara, Hedges, Larry V, & Konstantopoulos, Spyros. (2000). The effects of small classes on academic achievement: the results of the Tennessee class size experiment. American Educational Research Journal, 37(1), 123151. Petitta, Laura, Jiang, Lixin, & Palange, Miriam. (2015). The differential mediating roles of task, relations, and emotions collective efficacy on the link between dominance and performance: A multilevel study in sport teams. [Article]. Group Dynamics, 19(3), 181-199. Pion, Johan, Lenoir, Matthieu, Vandorpe, Barbara, & Segers, Veerle. (2015). Talent in female gymnastics: a survival analysis based upon performance characteristics. International Journal of Sports Medicine, 36(11), 935-940. Ross, W D, & Marfell-Jones, Michael J. (1991). Kinanthropometry. In MacDougall, J. Duncan, Wenger, Howard A. & Green, Howard J. (Eds.), Physiological testing of the highperformance athlete (pp. 223-308). Champaign, Ill Human Kinetics. Sands, Willian A. (2011). Fitness model of high-level gymnasts. In Jemni, Monem (Ed.), The science of gymnastics (pp. 22-25). London: Routledge. USA-Gymnastics. (2010). TOPS History-National TOP Averages and Standard Deviation/National Best Retrieved May 2th, 2016, from https://usagym.org/PDFs/Women/TOPs/Hi story/10topavg.pdf USA-Gymnastics. (2014). Talent Opportunity Program - testing manual Retrieved 24th september, 2014, from https://usagym.org/pages/women/video/top s.html Vandorpe, Barbara, Vandendriessche, Joric, Vaeyens, Roel, Pion, Johan, Lefevre, Johan, Philippaerts, Renaat, et al. (2011). Factors discriminating gymnasts by competitive level. International Journal of Sports Medicine, 32(8), 591-597. Corresponding author: Sarita Bacciotti Physical Education Course Federal University of Mato Grosso do Sul Campo Grande-MS-Brazil. E-mail: saritabacciotti@hotmail.com Science of Gymnastics Journal 139 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 THE EFFECT OF 6-WEEKS WHOLE BODY VIBRATION ON MUSCULAR PERFORMANCE ON YOUNG NONCOMPETITIVE FEMALE ARTISTIC GYMNASTS George Dallas1, Alkmini Savvathi1, Konstantinos Dallas2, Maria Maridaki1 1 National and Kapodistrian University of Athens - School of Physical Education and Sport Science, Athens, Greece 2 Athens, Greece _Original article Abstract The purpose of this study was to investigate the effects of a 6-week whole body vibrationintervention on muscle performance and flexibility on gymnasts. Twenty-two young non-competitive-moderate trained gymnasts that volunteered to participate in the study separated into either the vibration group or the no vibration group according to their training regime. The vibration intervention consisted of a 6-week whole-body vibration, 3 times per week and involved eccentric and concentric squatting movements on a vibration platform with the participants performing three exercises on the vibration device whereas for the no vibration group vibration platform was turned off. Five performance tests (20m running speed, sit & reach test, squat jump, counter movement jump and single leg squat (right leg and left leg) were performed at the beginning of the intervention, and after the end of 6-week intervention program. According to the results significant interaction effect between group and time was found for the running speed and Squat Jump test. On the contrary, significant main effect were found were found for time on the running speed, Squat Jump, Counter movement jump and single leg squat. Conclusively, it has been reported that Whole body vibration is an effective method to improve Squat Jump performance in young non-competitive female artistic gymnasts. Keywords: explosive strength, intervention, gymnastics. INTRODUCTION Whole Body Vibration (WBV) exercise or vibration training is a new type of exercise that has been reported to be an effective method to improve athletic performance (Cole & Mahoney, 2010). It is considered an ergogenic aid for training and competition (Bazett-Jones, Finch, & Dugan, 2008) and is potentially a less-consuming method for increasing power output than traditional training (Marin & Rhea, 2010). The main argument for using vibration for muscle training has been based on the assumption that strength improvements can be easily achieved during a short time (Dallas et al., 2015; Tsopani et al., 2014). When a person stands on a vibration platform it generates vertical sinusoidal vibration which are transmitted directly to the tendon, which in turn leads to activation of the alpha-motoneurons and initiates muscle contractions comparable to the "tonic Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 vibration reflex (TVR)" (Nortlund & Thorstensson, 2007). The transmission of mechanical oscillations from the vibrating platform may lead to physiological changes in muscle spindles, joint mechanoreceptors, higher brain activity and strength and power properties (Moezy, Olyaei, Hadian, Razi, Mohammad, & Faghihzadeh, 2008). TVR is the sustained contraction of a muscle due to the effect of vibration that activates muscle spindles, which are muscle receptors sensitive to stretch in the muscle. Afferent fibers send a signal to the spinal cord from the muscles spindles activating a reflex which causes the muscle to contract. TVR also causes an increase in recruitment and synchronization of motor units within the muscle via activation of muscle spindles and polysynaptic pathways (de Gail, Lance, & Neilson, 1966), which is seen as a temporary increase in the muscle activity. Furthermore, the increase in strength following WBV training may be due to elevated essential hormones, i.e. testosterone, growth hormone, and insulinlike growth factor. According to Rothmuller and Cafarelli (1995) vibration enhances the stretch reflex loop through the activation of the primary endings of the muscle spindle, which influences agonist muscle contraction while antagonists are simultaneously inhibited. Further, according to Cardinale & Bosco (2003), the acute enhancement of neuromuscular performance after vibration is probably related to an increase in the sensitivity of the stretch reflex. Furthermore, vibration appears to inhibit activation of antagonist muscles through Ia-inhibitory neurons, thus altering the intramuscular coordination patterns leading to a decreased braking force around the joints stimulated by vibration. Vibration might raise the muscle temperature due to the friction between the vibrating tissues, increase the blood flow, which could in turn enhance the extensibility of the muscle and ROM and change the pain threshold (Sands et al, 2008). Several studies showed that WBV training resulted in improved muscle strength or muscle performance. Specifically, it has been shown that exposure to WBV increases the explosive strength of the lower limbs, flexibility with or without stretching (Dallas & Kirialanis, 2013; Dallas, Kirialanis, & Mellos, 2014a; b; Dallas et al., 2015; Tsopani et al, 2014) and sprint running (Paradisis & Zacharogiannis, 2007). The majority of these studies are referred to female and male participants aged between 22-43 years (Carson, Popple, Verschueren, & Riek, 2010) or young adults (Chen, Liu, Chuang, Chung, & Shiang, 2014) or trained individuals; volleyball and beach volley athletes (Perez-Turpin et al., 2014) or female basketball athletes (Fernadez-Rio, Terrados, & Suman, 2012), whereas there is a strong to moderate evidence that long-term whole body vibration exercise can have positive effects regarding leg muscular performance (Annino et al., 2007; Chen et al., 2014; Fernandez-Rio et al., 2012; Dallas et al, 2017; Perez-Turpin et al., 2014; Preatoni et al., 2012). However, there are controversial results with regards to the effect WBV on explosive strength of lower limbs with some of these studies to indicate positive effect (Fernadez-Rio et al., 2012; Perez-Turpin et al., 2014; Preatoni et al, 2012), whereas other studies found no positive effect (Carson et al., 2010). In addition, a number of these studies have been applied a single bout of vibration to examine acute effect of vibration (Dallas & Kirialanis, 2013; Dallas, Kirialanis, & Mellos, 2014a; b; Tsopani et al, 2014), whereas other studies examine the effect of intervention program on muscle performance (Fernandez-Rio et al, 2012; Perez-Turpin et al, 2014; Preatoni et al, 2012; Sands et al, 2006). In addition, studies that are referred to gymnastics sports have examined mainly the vibration effect on flexibility in competitive gymnasts (Dallas & Kirialanis, 2013), on young gymnasts (Dallas et al., 2014a) or Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 high level gymnasts (Sands et al, 2006; Tsopani et al., 2014). Studies on long-term effects of vibration training have also yielded contradictory outcomes. Significant enhancement was found in the parameters evaluated by Annino et al. (2007), whereas other studies found no significant changes (Hand, Verschuere, & Osternig, 2009). More specifically, Paradisis and Zacharogiannis (2007) and Wyon, Guinan, and Hawkey (2010) concluded that the 6 weeks program improves sprint running and explosive power of lower limbs of healthy subjects. Perez-Turpin et al. (2014) studied 23 beach volleyball and volleyball athletes reported that implementation of 6-week WBV training increases leg strength more and leads to greater improvement in jump performance than traditional strength training. Different time protocols were also studied. Carson et al. (2010) compared a 4-week WBV training program to that without vibration. The subjects were aged 22-43 years and concluded that the non-vibration group increased the CMJ height; however, there were no significant improvements between groups. An 8 week WBV program by Chen et al. (2014) on injury-free young adults with average age of 20.3 years revealed a significant jumping improvement from 6.4 to 11.9%. Another 8 week training program (periodized) was carried out by Preatoni et al. (2012) on female softball and soccer athletes and concluded that combining conventional strength training with WBV doesn't improve the athletes' strength. The 8-week WBV training program of 22 ballerinas by Annino et al. (2007) concluded that it improves explosive strength on knee-extensor muscles although this is only short term. When comparing with and without vibration 14-week training programs of 31 female basketball athletes Fernandez-Rio et al. (2012) concluded that there were no significant improvements between the two programs even though there was strength increase. Further, a meta-analysis found that WBV has small and inconsistent acute and chronic effects on athletic performance in competitive and /or elite athletes (Hortobagyi, Lesisnski, Fernandez-del-Olmo, & Granacher, 2015). Although, there is a lack of data concerning the improvement of muscular strength in young gymnasts, previous data by Ronnestad (2004) support that vibration may enhance measures of explosiveness. Furthermore, as Kinser and colleagues (2008) stated the combination of vibration and stretching can enhance flexibility while maintaining explosive strength in young gymnasts aged 11.3 ± 2.6 years. In addition, Rohmert et al. (1989) showed that muscles with increased muscle length or tension are most affected by vibration. Therefore, it may be advantageous to use a combination of vibration and stretching as part of the warm-up for gymnasts, thus enhancing flexibility and maintaining or improving explosive performance (Stone et al, 2006). However, despite the above noted effect of vibration devises among subjects, scientific evidence on the efficacy of long term intervention vibration program on young gymnasts is lacking. Further, to our best knowledge, there are not scientific data related to the effect of long term intervention program on gymnasts. Therefore, the purpose of this study was to investigate the effect of a 6-week whole body vibration intervention on explosive strength of lower limbs, flexibility and running speed (RS) of young moderately trained artistic gymnasts. It was hypothesized that 8-week vibration intervention program will be more effective in improving muscular performance compared to the same program without vibration. METHODS Twenty-two young, healthy volunteers participated in this study (age 9.70 ± 0.95 years, body mass 34.47 ± 6.94 kg, and body height 137.67 ± 8.14 cm). All individuals, the last 2 years, were Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 participated in a scheduled training program for 3 times per week in clubs of artistic gymnastics and had no previous experience in WBV. They were randomly assigned to two groups, which included a WBV group (VG: n = 12) and a nonvibration group (NVG: n = 10). During the six-week intervention period, both groups were advised to continue with their regular habitual gymnastics training. Furthermore, participants incorporated an additional conditioning program at the end of each training session performing three exercises according to their protocol. Subjects were informed extensively about the experiment procedures and the possible risks or benefits of the project. They had no musculoskeletal injuries in the previous 6 months, and all parents provided written informed consent before participation of their children in the experimental design. The study was approved by the local institutional Review Board and all procedures were in accordance with the Helsinki declaration of 1975 as revised in 1996.Three days prior to the study, the anthropometric characteristics of subjects (age, body mass, body height) were measured and a familiarization session to get acquainted with the proper technique for the execution of the examined exercises (running speed, flexibility test, jumping tests) were performed. The vibration protocol consisted of a 6-week WBV training, which will be discussed in detailed herein. The intervention program was implemented at the end of each training session, as the technical skills training module is usually preceded. The total duration of training lasts about 60 minutes, with the most time, other than that of warm-up, devoted to the learning of simple technical exercises. Therefore, the degree of fatigue may be reduced to a minimum. On each session, each subject performed a 3-min standardized warm-up that include running at low intensity and light callisthenic exercises. Participants in the VG were exposed to vertical sinusoidal mechanical WBV while standing on the commercially available Power Plate® Next Generation WVB platform (Power Plate North America, Northbrook, Illinois), whereas participants in NVG performed the same exercises but the WBV platform was turned off. The vibration was set at 30 Hz, which produced a peak-to-peak amplitude of 2.5 mm and an acceleration of 2.28 g. The participants were exposed for 6 weeks using different execution forms of three exercises. Every training session was supervised by the researchers. In first exercise, participants were instructed to move through the eccentric portion of the movement for two sec and the concentric portion for two sec synchronized by a metronome operating with 20 bpm to a depth of approximately 90° of knee flexion. In exercise two and three, the participants standing on one leg, flexed their knee to a depth of approximately 90° of knee flexion and instructed to move as in the first exercise. The duration of 30 sec was used in hopes to improving the performance enhancement found by Cormie et al. (2006). During all the vibration-training session, the participants wore the same gymnastics shoes to avoid bruises and to standardize the damping of the vibration caused by the footwear. The training load on the vibration platform for the two groups, for each one exercise was as follows: Table 1 Training load for intervention protocol Week Series Duration (sec) 1-3 1 30 4 - 6_2_30_ The rest between the sets and exercises was 30 sec to provide a proper time for relaxation. As there are no scientific-based WBV programs the training program in the present study was based on similar protocols that resulted in significant changes in muscle performance (Torvinen et al, 2002a). A battery of tests (Running speed 20 m: RS; Flexibility test: S & R test; Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 Jumping tests) was performed at the start baseline (pre-test), one day after the end of sixth week (Post 1) and 5-min (Post 5) and 10-min (Post 10) after the end of the intervention protocol to measure the effects of training. The participants were informed about the test procedures and were asked to perform all these tests at maximum intensity. Before each test, the participants had one uninventive familiarization trial. Testing for each subject was completed in the same order during each testing period. All subjects participated in supervised training three days per week during the six-week training sessions. For all test items the Interclass reliability coefficients was estimated to be 0.91 - 0.96. In the first day of performance testing, the participants, after completion of the standardized 3-min warm-up, performed two maximal 20 m RS with subjects started from a crouched position. The sprint was performed in a gymnastic hall at a constant temperature of 22° C. The time was obtained using the Brower timing systems (Brower, USA). A rest period of 5 minutes was given between each trial. Flexibility was measured using the sit and reach test using a Flex-Tester box (Cranlea, UK). Participant, sitting barefoot on the floor with legs out straight ahead, were instructed to lean forward slowly as far as possible, toward a graduated ruler held on the box from -25 to +25, without bending their knees and held at the greatest stretch for 2 sec. The investigator has to be sure that there are no jerky movement on the part of the participant and that their fingertips remain at the same level and the legs flat. The score is recorded as the distance before or beyond the toes. The test was repeated twice with a rest period of 10 sec (15), and the best score was recorded. Jumping performance was measured using the squat jump (SJ), the counter movement jump (CMJ) and single leg squat (right leg (RL) and left leg (LL). Vertical jump tests were conducted on a switch mat connected to a digital timer (accuracy±0.001s, Ergojump, Psion XP, MA.GI.CA. Rome, Italy), which recorded the flight time (tf) of each single jump. The height of rise of the centre of mass in all jump tests was determined by the flight time and used in order to analyze the explosive strength characteristics of the leg muscles as reported elsewhere (Bosco et al., 1998). Prior to testing, the participants underwent one or two familiarization trials to ensure the proper performance technique for these three different jumps. The squat jump (SJ) started from a semi-squatting position with the knee flexed approximately at 90° that was maintained for 2 sec before jumping vertically. During SJ, the participants kept their trunk in an upright position and their hands on hips. The participants were instructed to perform two maximal trials with a rest period of 30 sec and the best jump was considered for further statistical analysis. For the counter movement jump (CMJ) the participants were instructed to perform a maximal vertical jump starting from upright position, with hands positioned at the hips to assess the lower-limb explosive performance capacity. The same regime as previous for SJ was followed. During CMJ, the participants kept their trunk in an upright position and their hands on hips. The participants were instructed to perform two maximal trials with a rest period of 30 sec and the best jump was considered for further statistical analysis. SJ involved the participants assuming a 90° knee bend position holding for two seconds, and jumping. The CMJ began in an upright position, had no pause, and was one, fluid, jumping movement. Depth for the eccentric portion of the CMJ was self-selected. The single leg squat (right and left), was performing under instructions that were given for the SJ test. The SPSS version 24 was used for the statistical analysis. The arithmetic mean, standard deviation, and range were calculated for each variable and trial. To explore the impact of time (pre, post1, post5, post10) and group (VG, NVG) on Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 the dependent variables, a two-way (group x time) ANOVA with repeated measures on the second factor was used for the statistical analysis. Sphericity was checked using Mauchly's test, and the Greenhouse-Geisser's correction on degrees of freedom was applied when necessary. Levene's test of equality of error variances was used to check the assumption of homogeneity of variances. In cases where interaction between time and group was detected, the simple effects were investigated, and Bonferonni's correction was used. In the absence of interaction, the main effects of the two factors (time and group) on the dependent variables were investigated. All statistical significances were tested at a = 0.05. RESULTS The statistical analyses revealed that the interaction effect between time and group was statistically insignificant for RS (F (3, 60) = 0.947, p > 0.05). On the contrary, significant main effect was found for time (F (3, 60) = 3.5 5 8, p < 0.019, n2 = 0,151), but the post hoc analysis revealed no significant differences among the factor's levels. Furthermore, no significant main effect was found for group (F (1, 20) = 4.222, p > 0.05). The mean values of the examined parameters are presented in table 1. Regarding S & R, the statistical analysis demonstrated no significant interaction effect between the two factors (F (3, 60) = 0.070, p > 0.05) and also no significant main effect for time (F (3, 60) = 0.777, p > 0.05) and group (F(1, 20) = 0.112, p > 0.05). The SJ results indicated a significant interaction effect between the two factors (F (3, 60) = 3.911, p < 0.042, n2 = 0.164). The post hoc analysis showed that squat jump significantly increased after 6 weeks vibration training, either it was measured immediately after the vibration (Mean difference = 3.417 cm, p < 0.05, 95% CI = 1.430-5.403 cm), or 5 (Mean difference = 3.417 cm, p < 0.05, 95% CI = 1.292-5.540 cm) and 10 min later (Mean difference = 3.000 cm, p < 0.05, 95% CI = - 1.1034.896 cm). There was no significant interaction effect between time and group for CMJ (F (3, 60) = 1.477, p > 0.05). While the analysis revealed a statistically significant main effect for time (F (3, 60) = 4.163, p < 0.024, n2 = 0.172), the post hoc analysis did not indicate any significant difference among the four levels. The main effect for group was not significant (F (1, 20) = 0.536, p > 0.05). Regarding the RL there was not found significant interaction (F (3, 60) = 0.471, p > 0.05) or main effect for group (F (1, 20) = 0.972, p > 0.05). On the contrary, it was found significant main effect for time (F (3, 60) = 4.918, p < 0.004, n2 = 0,197). The post hoc analysis did not indicate significant difference in the RL high among the four levels. As concerns the LL, the interaction effect between the two factors was found no significant (F (3, 60) = 1.143, p > 0.05). The main effect for group was also insignificant (F (1, 20) = 0.537, p > 0.05), but the main effect for time was found statistically significant (F (3, 60) = 5.3 3 4, p < 0.008). It was indicated by the post hoc analysis that the LL height was greater after 5 min of the vibration intervention, compared with the LL height measured before the 6 weeks vibration training. DISCUSION The primary findings of this study was that 6-week strength training produced significant improvement on VG in RS, and SJ performance of non-competitive young female artistic gymnasts, whereas nonsignificant improvement was found at all examined parameters on NVG. To date, quantity in WBV protocol and quantity to outcome relationship has not been established (Rehn, Lidstrom, Skoglund, & Lindstrom, 2007). Studies, with or without a successful outcome, use a number of different frequencies, durations and amplitudes using a progressive protocol thus increasing the intensity of the frequency and/or duration and/or Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 amplitude of the WBV protocol (Cole & Mahoney, 2010). Although non-significant main effect was found for group, VG showed greater improvement (4.41%) on running speed (RS) compared to NVG (1.64%) after the end of the vibration protocol. These results are in accordance with data of Paradisis and Zacharogiannis (2007) who concluded that there is a 2.10% improvement in the time of 60m sprinting. The non-significant differences between two groups in our study verify previous results of Roberts, Hunter, Hopkins, & Feland (2009) who state that there were no observed differences in the 30 m sprint times in collegiate athletes. Conversely, our results opposed to those of Cole and Mahoney (2010) who found that WBV training may have had a small detrimental effect on speed during a 40m-dash test. However, other data failed to provide evidence that acute WBV stimulus positively affects sprint performance (Cochrane, 2013; Roberts et al, 2009). It has been observed that WBV training induces tonic vibration reflection; a higher activation of the muscular spindles and motor neurons which decreases the Electromagnetic delay (EMD) and increases the motor units. The neurological theory of muscular coordination could be a reason for this fast improvement in performance; motor neurons in one practical group of muscles and joints are prepared and there are improvements: in motor units coordination and integration, in synergist muscles co-contraction, and in antagonist muscles inhibition (Cardinale & Wakeling, 2005). Table 2 Mean values and standard deviations on various measurements. Pre Post 0 VG NVG VG NVG RS (sec) 4.26 ± 0.17 4.32 ± 0.32 4.08 ± 0.26t 4.25 ± 0.22 S & R (cm) 28.00 ± 4.99 27.20 ± 4.73 27.67 ± 5.10 27.20 ± 4.80 SJ (cm) 18.75 ± 2.73 20.20 ± 3.01 22.17 ± 3.32t 21.00 ± 2.71 CMJ (cm) 21.69 ± 2.00 21.94 ± 2.73 23.74 ± 3.60t 22.66 ± 2.83 RL (cm) 10.41 ± 1.50 9.70 ± 3.16 12.17 ± 2.65t 10.80 ± 2.61 LL (cm) 9.41 ± 1.97 9.50 ± 2.27 11.16 ± 2.82t 10.10 ± 2.28 Post 5 Post 10 VG NVG VG NVG RS (sec) 3.97 ± 0.26t 4.22 ± 0.22 4.07 ± 0.27 4.26 ± 0.27 S & R (cm) 28.17 ± 5.00 27.70 ± 4.44 28.67 ± 4.83 27.80 ± 4.42 SJ (cm) 22.17 ± 3.40t 21.10 ± 3.14 21.75 ± 3.52t 21.20 ± 3.08 CMJ (cm) 23.81 ± 3.54t 22.42 ± 3.03 23.37 ± 3.22 22.11 ± 2.99 RL (cm) 12.25 ± 2.89t 10.90 ± 2.88 11.33 ± 2.46 10.80 ± 3.01 LL (cm) 11.33 ± 2.42t 10.20 ± 2.25 10.91 ± 2.64 10.20 ± 3.08 t denote significant difference compared to baseline (pre) values (p <.05) In Sit & Reach test (S & R), VG showed a slight mean decrease by -0.33 cm (1.21%) immediately after the end of 6-weeks vibration, whereas NVG remain unchangeable, that means WBV training may not be reflected in improvements in flexibility. This finding is in accordance with data of Cole and Mahoney (2010), which revealed no significant effect of WBV on flexibility of the hamstrings and lower back after 5-weeks of WBV training. However, our results opposed to those of Marshall and Wyon (2012) who found a significant large increase by 30% in ROM in young trained dancers without increasing thigh and calf circumferences Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 after a 4-week vibration protocol and those of Sands et al. (2006) who a significant improvement, after 4-week vibration training on young high trained male gymnasts, on one split side performance. However, the improvement by 4.17% that appeared in our study 10 min after the end of the vibration protocol in VG was greater compared to that of NVG (2.80%), which suggest that the vibration exposure may have activated the Ia inhibitory interneurones of the antagonist muscle. It is mentioned that these discrepancies on the aforementioned results may be attributed to the different vibration protocols that were applied, the status and the chronological age of the subjects. Results showed that a 6-week Whole Body Vibration induced, immediately after of the end of vibration protocol, a significant percentage improvement by 18.81% on squat jump (SJ) on VG which is significant greater from those of NVG (3.96%). The improvement that showed VG support previous findings which revealed that WBV increases explosive strength of lower limbs (Fernadez-Rio et al., 2012). Furthermore, the percentage improvements of VG 5 and 10 min after the end of vibration protocol (18.81% and 16.00%, respectively) are much greater from those of NVG (4.45% and 4.95%, respectively). In the present study the applied protocol of 6-weeks had a positive effect on counter movement jump (CMJ) performance. It was found that VG showed greater improvements (9.45%) than NVG (0.96%). This improvement by 9.45% support previous findings of Wyon et al. (2010) that found a beneficial effect of vertical jump height after 6-week vibration intervention in moderately trained undergraduate female dance students. Furthermore, results of our study are in congruence with those of other authors that founded an increase by 1.49% - 9.0% after several weeks of WBV training (Annino et al., 2007; Armstrong, Grinnell, & Warren, 2010; Bazett-Jones et al, 2008; Cole & Mahoney, 2010; Hand et al, 2009). Furthermore, the percentage improvements of VG 5 and 10 min after the end of vibration protocol (9.77% and 7.70%, respectively) are much greater from those of NVG (3.36% and 0.77%, respectively). Suggested neuromuscular improvement mechanisms are: increased corticomotor excitability and decreased short-interval intracortical inhibition, increased muscle activity due to dampening of the vibrational oscillations (Boyer & Nigg, 2007), increased motor unit activity (Pollock, Woledge, Martin, & Newham, 2012). Significant greater improvements were found in right leg (RL) and left leg (LL) on VG immediately after the end of vibration protocol (16.91% and 18.57%, respectively) compared to NVG (11.34% and 6.31%, respectively). These results confirm findings of Shin, Lee, & Song. (2015) who found a considerably larger increase on SLJ in unilateral vibratory stimulation group (UVSG: 21% for the weak leg [WL]) and 12% for the strong leg [SL]) in comparison to the NVG (1.86% and 1.98% for the WL and SL, respectively). Further, the bilateral deficit (BLD) of 4.41% found in our study, support data of Costa, Moreira, Cavalcanti, Krinski, and Aoki (2015) who reported a strength reduction of 11%, meaning a BLD, after a resistance training protocol that consisted of 3 sets of leg extensions using a load of 50% 1RM. The bilateral deficit is a phenomenon where the sum of force produced by each leg individually is greater than the force produced by both legs combined in bilateral movement and is due to neural inhibition during bilateral tasks. WBV induces a non-voluntary muscle contraction i.e. TVR, activating amotor neurons and increasing the sensitivity of primary ending of the muscle spindles thus stimulating the Ia afferent fiber of the muscle spindle. Moreover, more muscles are recruited via the muscle spindles and neuron bundles. According to Henneman's size principle, as contraction strength increases, motor units are Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 recruited from smallest to largest. According to Swearingen et al. (2011), small type I muscles are recruited before large type II, and with more large type II muscles, muscle strength and movement improve. According to the above, even slight WBV training can significantly affected muscle recruitment required for the single leg jump. Although knowledge the neurological and physiological mechanism of WBV training is limited there has been numerous research into the mechanisms through which WBV training affects performance. Our study sample was fairly small and thus the results of the WBV should be interpreted with caution. Furthermore, although there were no reported injuries during the period of our study, care should be taken vibration training is believed to be stressful for this particular age group. Additional studies are required with longer interventions in order to verify whether WBV training as a complement to resistance training produces specific benefits. The results of our study, however, may encourage trainers to include WBV sessions in their training programs so as to increase leg strength and jump height. Certain limitations do not allow the generalization of this study. The results refer to a specific category of athletes and a specific level of technical training. The intervention protocol was applied after the end of each workout, with the potential for the resulting fatigue to affect the end result. To carry out a study by applying the intervention protocol at the beginning of the training sessions and at athletes of competitive level would give different results. CONCLUSIONS The findings of this study demonstrate that the implementation of 6-week WBV training in non-competitive young female artistic gymnasts improve leg strength more and leads to greater improvement in jump performance than traditional strength training (NVG). REFERENCES Annino, G., Padua, E., Castagna, C., di Slavo, V., Minichella, S., Tsarpela, O., Manzi, V., & D'Ottavio, S. (2007). Effect of whole-body vibration training on lower limb performance in selected high-level ballet students. Journal of Strength and Conditioning Research, 21, 1072-1076. Armstrong, W.J., Grinnell, D.C., & Warren, G.S. (2010). The acute effect of whole-body vibration on the vertical jump height. Journal of Strength and Conditioning Research, 24, 2835-2839. Bazett-Jones, DM., Finch, H.W., & Dugan, E.L. (2008). Comparing the effects of various whole-body vibration accelerations on counter-movement jump performance. Journal of Sports Science and Medicine, 7, 144-150. Bosco, C., Cardinale, M., Tsarpela, O., Colli, R., Tihanyi, J., von Duvillard, S.P., & Viru, A. (1998). The influence of whole body vibration on jumping performance, Biology of Sport, 15, 157164. Boyer, K.A., & Nigg, M. (2007). Changes in muscle activity in response to different impact forces affect soft tissue compartment mechanical properties. Journal of Biomechanics and Engineering, 129, 594-602. Cardinale, M., & Bosco, C. (2003). The use of vibration as an exercise intervention. Exercise and Sport Science Review, 31, 3-7 Cardinale, M., & Wakeling, J. (2005). Whole body vibration exercise: are vibrations good for you? British Journal of Sports Medicine, 39, 585-589. Carson, R.G., Popple, A.E., Verschueren, S.M.P., & Riek, S. (2010). Superimposed vibration confers no additional benefit compared with resistance training alone. Scandinavian Journal of Medicine and Science in Sports, 20, 827-833. Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 Chen, C.H., Liu, C., Chuang, L.R., Chung, PH., & Shiang, T.Y. (2014). Chronic effects of whole-body vibration on jumping performance and body balance using different frequencies and amplitudes with identical acceleration load. Journal of Science and Medicine in Sport, 17, 107112. Cochrane, D.J. (2013). The effect of acute vibration exercise on short-distance sprinting and reactive agility. Journal of Sports Science and Medicine, 12, 497-501. Cole, K.J., & Mahoney, S.E. (2010). Effect of five weeks of Whole Body Vibration training on speed, power, and flexibility. Clinical Kinesiology, 64, 1-7. Cormie, P., Deane, R.S., Triplett, N.T., & McBride, J.M. (2006). Acute effects of whole-body vibration on muscle activity, strength, and power. Journal of Strength and Conditioning Research, 20(2), 257-261 Costa, E.C., Moreira, A., Cavalcanti, B., Krinski, K., & Aoki, M.S. (2015). Effect of unilateral and bilateral resistance exercise on maximal voluntary strength, total volume of load lifted, and perceptual and metabolic responses. Biology of Sport, 32, 35-40. Dallas, G., & Kirialanis, P. (2013). The effect of two different conditions of whole-body vibration on flexibility and jumping performance on artistic gymnasts. Science of Gymnastics Journal, 5, 67-77. Dallas, G., Kirialanis, P., and Mellos, V. (2014a). The acute effect of Whole-Body Vibration on flexibility and explosive strength of young gymnasts. Biology of Sport, 31, 233-237. Dallas, G., Paradisis, G., Kirialanis, P., Mellos, V., Argitaki, P., & Smirniotou A. (2015). The acute effect of different training load of whole body vibration on flexibility and explosive strength of lower limbs on divers. Biology of Sport, 32, 235241. Dallas, G., Tsiganos, G., Tsolakis, Ch., Tsopani, D., Di Cagno, A., & Smirniotou, A. (2014b). The acute effect of different stretching methods on flexibility and jumping performance in competitive artistic gymnasts. Journal of Sport Medicine and Physical Fittness, 54, 683-690. Dallas, G., Mavidis, A., Kirialanis, P., Papouliakos, S. (2017). The effect of 8-weeks of whole body vibration training on static balance and explosive strength of lower limbs on physical education students. Acta Gymnica, 47(4): 153-160. De Gail, P., Lance, J.W., & Neilson, P.D. (1966). Differential effects on tonic and phasic reflex mechanisms produced by vibration of muscles in man. Journal of Neurolology and Neurosurgery Psychiatry, 29, 1-11. Fernandez-Rio, J., Terrados, N., & Suman, O. (2012). Long-term effects of whole-body vibration training in high-level female basketball players. Journal of Sports Medicine and Physical Fitness, 52, 18-26. Hand, J., Verschuere, S., & Osternig, L. (2009). A Comparison of Whole-Body Vibration and Resistance Training on Total Work in the Rotator Cuff. Journal of Athletic Training, 44, 469-474. Hortobagyi, T., Lesisnski, M., Fernandez-del-Olmo, M, & Granacher, U. (2015). Small and inconsistent effects of whole body vibration on athletic performance: a systematic review and meta-analysis. European Journal of Applied Physiology, 115, 1605-1625. Marin, P.J., Rhea,& MR. (2010). Effects of Vibration Training on Muscle Strength: A Meta-Analysis. Journal of Strength and Conditioning Research, 24, 548-556. Marshall, L.C., & Wyon, M.A. (2012). The effect of whole-body vibration on jump height and active range of movement in female dancers. Journal of Strength and Conditioning Research, 26, 789-793. Moezy, A., Olyaei, G., Hadian, M., Razi, M., & Faghihzadeh, S. (2008). A comparative study of whole body vibration training and conventional training on knee proprioception and postural stability after Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 anterior cruciate ligament reconstruction. British Journal of Sport Medicine, 42, 373378 Nordlund, M.M., & Thorstensson, A. (2007). Strength training effects of whole body vibration? Scandinavian Journal of Medicine and Science in Sport,17, 12-17 Paradisis, G., & Zacharogiannis, E. (2007). Effects of whole-body vibration training on sprint running kinematics and explosive strength performance. Journal of Sport Science and Medicine, 6, 44-49. Perez-Turpin, J.A., Zmijewski, P., Jimenez-Olmedo, J.M., Jove-Tossi, M.A., Martinez-Carbonell, A., Suarez-Llorca, C., & Andreu-Cabrera, E. (2014). Effects of whole-body-vibration on strength and jumping performance in volleyball and each volleyball players. Biology of Sport, 31, 239-245. Pollock, R.D., Woledge, R.C., Martin, F.C., & Newham, D.J. (2012). Effects of whole body vibration on motor unit recruitment and threshold. Journal of Applied Physiology, 112, 388-395. Preatoni, E., Colombo, A., Verga, M., Galvani, C., Faina, M., Rodano, R., Preatoni, E., & Cardinale, M. (2012). The effects of whole body vibration in isolation or combined with strength training in female athletes. Journal of Strength and Conditioning Research, 26, 2495-2506. Rehn, B., Lidstrom, J., Skoglund, J., & Lindstrom, B. (2007). Effects on leg muscular performance from whole-body vibration exercise: A systematic review. Scandinavian Journal of Medicine and Science in Sports, 17, 2-11. Roberts, B., Hunter, I., Hopkins, T.Y., & Feland, B. (2009). The short-term effect of whole-body-vibration training on sprint start performance in collegiate athletes. International Journal of Exercise and Science, 2, 264-267. Rohmert, W., Wos, H., Norlander, S., & Helbig, R. (1989). Effects of vibration on arm and shoulder muscles in three body postures. European Journal of Applied Physiology and Occupation Physiology, 59, 243-248. Ronnestad, B.R. (2004). Comparing the performance-enhancing effects of squats on a vibration plate with conventional squats in recreational resistance-trained men. Journal of Strength and Conditioning Research, 18(4), 839845 Rothmuller, C., & Cafarelli, E. (1995). Effect of vibration on antagonist muscle coactivation during progressive fatigue in humans. Journal of Physiology, 485, 857-864 Sands, W.A., McNeal, J.R., Stone, M.H., Russell, E.M., & Jemni, M. (2006). Flexibility Enhancement with Vibration: Acute and Long-Term. Medicine and Science in Sports and Exercise, 38, 720-5 Sands, W.A., McNeal, J.R., Stone, M.H., Haff, G.G., & Kinser, A.M. (2008). Effect of vibration on forward split flexibility and pain perception in young male gymnasts. International Journal of Sports and Physiology Performance, 3, 469-81 Shin, S., Lee, K., & Song, C. (2015). Acute effects of unilateral whole body vibration training non single leg vertical jump height and symmetry in healthy men. Journal of Physical Therapy and Science, 27, 3923-3928. Stone, M.H., Ramsey, M.W., Kinser, A.M., O'Bryant, H.S., Ayres, C., & Sands, W.A. (2006). Stretching: acute and chronic? The potential consequences. Strength and Condition Journal, 28, 66-74. Swearingen, J., Lawrence, E., Stevens, J., Jackson, C., Waggy, C., & Davis, D.S. (2011). Correlation of single leg vertical jump, single leg hop for distance, and single leg hop for time. Physical Therapy in Sport, 12, 194-198. Torvinen, S., Kannus, P., Sievanen, H., et al. (2002a). Effect of four-month vertical whole body vibration on performance and balance. Medicine and Science in Sport and Exercise, 34(9), 1523-1528 Tsopani, D., Dallas, G., Tsiganos, G., Papouliakos, S., Di Cagno, A., Korres, G., Riga, M., & Korres, St. (2014). Short-term Science of Gymnastics Journal 151 Science of Gymnastics Journal Dallas G., Savvathi A., Dallas K., Maridaki M.: THE EFFECT OF 6-WEEKS WHOLE. Vol. 11 Issue 2: 151 - 162 effect of whole-body vibration training on balance, flexibility and lower limb explosive strength in elite rhythmic gymnasts. Human Movement Science, 33, 149-158. Wyon, M., Guinan, D., & Hawkey, A. (2010). Whole-body vibration training increases vertical jump height in a dance population. Journal of Strength and Conditioning Research, 24, 866-870. Corresponding author: George Dallas National and Kapodistrian University of Athens - Physical Education and Sport Scienc 41, Ethnikis Antistaseos Dafne, Athens 17237, Greece T: +0030 210 66 43 704 F: ++0030 210 727 6028 E-mail: gdallas@phed.uoa.gr Science of Gymnastics Journal 151 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 MECHANICAL LOWER LIMB MUSCLE FUNCTION AND ITS ASSOCIATION WITH PERFORMANCE IN ELITE TEAM GYMNASTS Ole H. Hansen 1, Lars G. Hvid 2, Per Aagaard1 and Kurt Jensen 1 department of Sports Science and Clinical Biomechanics, University of Southern Denmark, Odense, Denmark 2Section of Sport Science, Department of Public Health, Aarhus University, Aarhus, Denmark _Original article Abstract TeamGym (TG) differs from individual gymnastics as it is performed in teams including 6-12 participants competing in acrobatic performance in three disciplines: trampette jumping, tumbling track jumping, and floor exercises. The physical demands required by TG athletes largely remain unknown, and likely are dictated by the specific disciplines and equipment used. This study aimed at describing physiological capacity by investigating mechanical lower limb muscle function and its association with TG performance in 24 senior elite (12 males, 12 females) team gymnasts. Methods: Anthropometrical data as well as 25m sprint ability, repetitive jumps (RJ), countermovement jumping (CMJ), drop jumping from a height of 48cm (DJ48), maximal isometric leg press muscle strength (MVC) and rate of force development (RFD) were measured. Results: Significant sex differences (p<0.05) were observed for all variables, except MVC. Total sprint times were 3.36±0.1s in males vs. 3.70±0.1s in females, CMJ height 0.51±0.05 vs. 0.41±0.03m, DJ48 rebound height 0.43±0.06 vs. 0.34±0.06m, with no difference in concentric peak power production between CMJ and DJ48. MVC was 38.3±9.9N/kg in males vs. 36.4±9.2N/kg in females. In female gymnasts, correlations (r2=0.41-0.46, p<0.05) were found between trampette and tumbling performance and sprint ability. In male gymnasts, correlations (r2=0.44, p<0.05) emerged between trampette performance and relative RFD (%MVC/s). Conclusions: Moderate associations were found between mechanical lower limb muscle function and functional tumbling performance in male and female TeamGym, indicating that performance in elite TeamGym also relies on factors other than isolated mechanical muscle function. Keywords: TeamGym, sprint, jumping, maximal strength, RFD, acrobatic performance. INTRODUCTION The sport of gymnastics comprises different disciplines like sports acrobatics, rhythmic sports gymnastics and artistic gymnastics. As a sub-discipline, TeamGym (TG) is becoming increasingly popular with European Championships conducted biannually since 1993 (Elbaek & Froberg 1993). In TG the gymnasts compete in teams of 6-12 gymnasts in 3 disciplines: 1) advanced acrobatics on the trampette, with and without vaulting table, 2) tumbling with both forward, and Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 backwards acrobatic routines, and 3) floor routine with dance elements, acrobatics, and choreography. The performances of the gymnasts are evaluated on basis of the difficulty and the quality of the acrobatic drills and the floor routine (Sjöstrand et al. 2015). The combined scores assigned to the 3 disciplines determine the final ranking of the teams. Few scientific studies have investigated TG, most of them with a focus on common types and frequency of musculoskeletal injuries (Harringe et al. 2004; Harringe et al. 2007). In general, the acrobatic performance of gymnastics involves highly complex full-body movements and TG has been shown to induce substantial stress on the lower extremities (Harringe et al. 2004; Harringe et al. 2007; Lund & Myklebust 2011). In an early study conducted in 1992, the year prior to the first European Championship, Elbaek & Froberg (1993) investigated the capacity of Danish clublevel TG gymnasts. They concluded that TG athletes were characterized by superior anaerobic power and strength levels compared to a control group of physical education students. Since this early report, exercise equipment such as the tumbling track and the trampette has evolved considerably worldwide resulting in increased levels of difficulty of the acrobatic skills. Accordingly, an increased demand for high technical and physical capacity has been introduced in modern TG competitions. Power generation, explosive-type movements and high anaerobic capacity are among performance-decisive factors across various gymnastic disciplines. (Elbaek & Froberg 1993; Bale & Goodway 1990; Jensen et al. 2013; Jemni et al. 2000; Suchomel et al. 2016; Aleksic-Veljkovic et al. 2016). However, these aspects have not previously been systematically examined in elite TG athletes. Therefore, the purpose of the present study was to describe the physiological capacity of international level TG athletes by examining mechanical lower limb muscle function (Power, MVC strength, RFD) and its associations to TG performance in Danish elite male and female team gymnasts. It was hypothesised that several of the test outcome would be related to TG performance among these athletes. METHODS The participants recruited for the study were senior gymnasts from the 2016 National Danish Teams (male, female and mix). Testing was performed prior to the Danish National Championship. Exclusion criterion was current musculoskeletal injuries to the back or the lower extremities. Participants were experienced in performing regular testing in their clubs or at National Team training camps. They gave their informed consent to participate in the test protocols and were well informed of the potential risks. A total of 12 male and 12 female team gymnasts completed the tests. Age, body height, body weight, body fat, training hours and TG performance are reported in Table 1. Participants reported their current performance level in different acrobatically skills. The reported skills were divided into (i) trampette jumping and (ii) tumbling jumping. The international Code of Points (Sjostrand et al. 2015) were used to rate the difficulty of performance (D-score) and the total score of the three most difficult skills in both trampette and tumbling were selected for analysis. Notably, the D-score has no upper limits. To the best knowledge of the authors, the highest scores for an individual skill in trampette or series of skills in tumbling recorded at any international championships (males) have been 2.35 and 2.9 points, respectively. All participants were instructed to refrain from strenuous exercise 24 hours prior to testing. Participants received instructional videos and photos of the tests prior to testing, alongside written information about the test procedures. The tests were executed in the presented order. Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 All participants performed the tests barefooted wearing regular training clothes. Body height was measured using a standard wall measuring scale and the nearest 0.5cm was registered (SECA, Hamburg, Germany). Body and fat mass were measured using bioelectrical impedance (Tanita MC-780MA Body Composition Analyzer, Tokyo, Japan). Time to sprint 25m from a standing start with intermediate times recorded at 5, 10, 20m and 25m was assessed using an automatic photo cell-based timing system (Swift Performance SpeedLight Dual-beam, Wacol, Australia). The start of the sprint was initiated from a sensor placed on the ground at the 0m marking. Participnts warmed up for 5 minutes by graded running and were instructed to complete two sprints of ~80-90% of their maximal voluntary effort. Participants performed the test barefooted on a wooden floor. Participants performed one submaximal trial (~80-90% of maximal effort) running between the light gates to get familiarized with the equipment. Subsequently, participants performed three maximal trials, where they were instructed to pass the 25m markers as fast as possible. In case of continous improvement over the three trials, an extra trial was given. The test leader (OHH) provided strong verbal encouragement during all trials. RJ was performed on a contact-mat (0.6 by 0.6m) (Swift Performance SpeedMat, Wacon, Australia) as previously described (Herbert-Losier & Eriksson 2014). The test consisted of a series of jumps where the aim was to perform 10 continuous vertical jumps as high as possible using the shortest possible contact time. The test was performed with the hands placed on the hips and participants performed as many submaximal trials as desired before performing two maximal trials. First and final jump were excluded, leaving 8 jumps for analysis. Reactive strength index (RSI = jump height / contact time) was calculated (Young 1995). The trial with the highest RSI was selected for further analysis. CMJ was performed on a force plate embedded into the floor of the Lab (0.6 by 0.4m) (Kistler 9281 B, Amherst, US) as described in detail previously (Thorlund et al. 2008). The vertical ground reaction force signal (Fz) was digitally sampled at 1000Hz. Participants stood still on the force plate with their hands placed at their hips and then performed a smooth movement into a self-chosen depth, followed by a fast-upward movement resulting in take-off and a vertical flight phase (cf. Fig.2 in Thorlund et al. 2008). The participants performed a self-chosen number of familiarization trials maybe at ~80% of maximal effort before completing three CMJ trials at maximal effort. The trial with the highest jump was selected for analyses. Maximal vertical jump height of the Body Centre of Mass (BCM) and peak leg extensor power during the concentric take-off phase were calculated by second order integration of the Fz signal as described in detail elsewhere (Thorlund et al. 2008). DJ48 was performed from a 0.48m elevated drop force plate (0.51 by 0.46m) (AMTI R6-1000, Watertown, MA) while landing on the Kistler force plate used for CMJ. The Fz signals from both the AMTI and Kistler force plates were synchronously digitally sampled at 1000Hz. The analytical approach regarding the dual force plate method has previously been described in details by Baca (1999). The Kistler force plate acted as the rebound surface, whereupon the participants performed a drop landing followed by a rapid rebound movement (CMJ) followed by a rebound flight phase and subsequently landing on the ground level force plate (Kistler). Maximal vertical BCM flight height and peak leg extensor power during the concentric rebound phase were calculated as described above for the CMJ. Participants performed a self-chosen number of Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 familiarization trials at ~80% of maximal effort and subsequently performed a minimum of three trials at maximal effort. Maximal strength and rate of force development (RFD) of the participants' dominant leg were assessed using static leg press testing in a custom-built leg press device with a fixed footplate instrumented with piezoelectric force transducers (Kistler 9367/8 B, Amherst, US) (Caserotti et al. 2008). The force signal was digitally sampled at 1000Hz, and subsequently lowpass filtered by a digital fourth-order, zero-lag Butterworth filter using a cut-off frequency of 10Hz. The latter was carried out off-line using a custom-built MatLab macro (MatLab 15 a. MathWorks, MA, US). Participants were seated with their thigh in a horizontal position and the knee joint fixed at a 120° position (0° = full extension). Arms were crossed over the chest and the back kept straight during all trials. Following three submaximal warm-up trials, three trials were performed at maximal voluntary effort. Participants were instructed to produce as much force as possible and to reach maximal force output as fast as possible. Participants received real-time visual feedback of the force output and were verbally encouraged during all trials. If there were a variation of more than 5% between the two best trials, an additional trial was performed. The trial with the highest force output (maximal voluntary contraction, MVC) was selected for further analysis. Onset of contraction was defined as the instant when force production exceeded the baseline level force by 2% of the maximal force value. All trials with a visible initial countermovement were discarded from the analysis. Contractile rate of force development (RFD) was derived as the average tangential slope of the force-time curve (Aforce/Atime) calculated in the time interval 0-30ms and 0-100ms relative to the onset of contraction (RFD30ms and RFD100ms). Additionally, relative RFD was calculated as RFD30ms and RFD100ms normalized to MVC to examine qualitative muscle properties during the initial phase of rising muscle force (Aagaard et al. 2002). Statistical analysis was performed using SPSS 24.0. Sex difference was tested with an unpaired t-test along with a 95% confidence interval of the difference of means. Assumption of normality was tested using the Shapiro-Wilk test for each variable and Pearson's product moment correlation was used to examine potential relationships between selected tests variables and TG performance. Results are presented as mean ± standard deviation (SD) unless otherwise stated. RESULTS Differences between male and female TG athletes were noted for all variables. Males were 11% older, 8% taller, had a 19% greater body mass and 8% lower fat mass than the females. Further, male athletes demonstrated ~50% greater acrobatic difficulty scores than female athletes, despite no significant difference in overall training volume (Table 1). During 25m sprint testing male TG athletes demonstrated 6-11% faster split-intervals along with 9.2% faster total sprint time compared to female athletes (Table 2). During RJ testing, males had 48% higher RSI than female participants. During stretch-shortening cycle (SSC) testing male TG athletes jumped 24% higher in CMJ accompanied by a 24% greater peak power per kg body weight and jumped 25% higher in the DJ48 accompanied by a 18% greater peak power per kg body weight compared to female athletes (Table 2). No sex differences emerged for MVC expressed relative to body mass or for absolute and relative RFD obtained during leg press testing (Table 2). One male participant was excluded in the MVC analysis due to corrupted data signals. Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 Table 1 Anthropometry and TG performance including P values and 95% confidence intervals of the difference between males and females. Difference of Male Female P mean 95% CI Age (years) 22.2 ± 2.3 20.0 ± 2.3 = 0.02 (0.3, 4.1) Height (cm) 175.8 ± 5.5 163.0 ± 5.7 <0.001 (8.4, 17.3) Body mass (kg) 73.3 ± 8.3 61.4 ± 5.8 <0.001 (6.1, 17.6) Fat mass (%) 13.1 ± 3.0 21.0 ± 3.6 <0.001 (-10.5, 5.2) Training volume (hours) 9.9 ± 2.6 9.7 ± 1.6 >0.05 (-1.5, 1.5) Tumbling performance (p) 5.8 ± 0.6 3.9 ± 0.5 <0.001 (1.4, 2.4) Trampette performance (p) 5.5 ± 0.4 3.7 ± 0.3 <0.001 (1.5, 2.1) Table 2 Sprint performance and lower limb mechanical muscle function in male and female elite team gymnasts including 95% confidence intervals of the difference between males and females. Male Female P Difference of means 95% CI 25 meters sprint test Sprint 5m (s) 0.84 ± 0.03 0.92 ± 0.04 <0 .001 (-0.1, -0.04) Sprint 10m (s) 1.57 ± 0.06 1.72 ± 0.05 <0 .001 (-0.2, -0.1) Sprint 20m (s) 2.79 ± 0.09 3.06 ± 0.07 <0 .001 (-0.3, -0.2) Sprint 25m (s) 3.36 ± 0.09 3.70 ± 0.09 <0 .001 (-0.4, -0.3) RJ RSI (height/contact-time) 195 ± 32 132 ± 36 <0 .001 (36.1, 90.5) CMJ Jump height (cm) 50.7 ± 4.5 41.0 ± 3.0 <0 .001 (6.5, 12.9) Peak power (W/kg) 58.9 ± 7.9 47.6 ± 6.1 <0 .001 (5.7, 16.9) DJ48 Jump height (cm) 42.7 ± 5.8 34.2 ± 6.1 <0 .001 (3.5, 13.5) Peak power (W/kg) 57.1 ± 4.0 48.3 ± 6.4 <0 .001 (4.5, 13.0) Isometric leg press test Peak MVC (N/kg) 38.3 ± 9.9 36 ± 9 >0 05 (-5.9, 9.8) RFD30ms (N/kg/s) 96.3 ± 27.3 86± 43 >0. 05 (-18.7, 40.0) RFD100ms (N/kg/s) 117.3 ± 38,9 110 ± 56 >0 05 (-31.5, 46.6) Relative RFD30ms (%MVC/s) 253 ± 51 230 ± 79 >0 05 (-31.0, 76.9) Relative RFD^ms (%MVC/s) 309.0 ± 88.5 298± 112 >0 05 (-70.7, 93.69 Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 Table 3 Correlation analysis of male and female team gymnasts' test variables and acrobatic performance. Male Female Total tumbling Total trampette Total tumbling Total trampette performance performance performance performance Pearson P- Pearson P- Pearson P- Pearson P- Variables Correlation value Correlation value Correlation value Correlation value Sprint 5m (s) -0.065 0.842 -0.031 0.924 -0.622 0.031 -0.638 0.026 Sprint 10m (s) -0.001 0.997 0.023 0.942 -0.515 0.087 -0.672 0.017 Sprint 20m (s) 0.143 0.657 -0.010 0.975 -0.431 0.162 -0.680 0.015 Sprint 25m (s) 0.221 0.489 -0.040 0.902 -0.377 0.227 -0.672 0.017 RJ (RSI) 0.113 0.726 -0.055 0.866 0.367 0.240 0.164 0.612 CMJ (cm) -0.212 0.509 -0.105 0.745 0.099 0.759 0.416 0.178 CMJ (W/kg) -0.314 0.320 -0.162 0.615 0.245 0.442 0.430 0.163 DJ48 (cm) -0.417 0.177 0.162 0.615 0.054 0.867 0.451 0.141 DJ48 (W/kg) -0.260 0.414 -0.031 0.923 0.204 0.525 0.515 0.087 Peak MVC (N) -0.210 0.536 -0.348 0.295 0.416 0.179 -0.042 0.898 Peak force (N/kg) -0.140 0.682 -0.342 0.303 0.408 0.188 -0.144 0.655 RFD30ms (N/kg/s) -0.198 0.560 0.088 0.798 0.476 0.118 0.125 0.698 RFDi00ms(N//kg/s) -0.189 0.579 0.102 0.765 0.444 0.148 0.128 0.693 Relative RFD30ms -0.097 0.778 0.663 0.026 0.245 0.444 0.189 0.555 Relative RFDi00ms -0.026 0.939 0.440 0.175 0.174 0.588 0.146 0.651 Group means ± SD. Abbreviation: RJ: Reactive Jumps; RSI: Reactive Strength Index, measured by a series of 8 jumps as the [mean jump height / mean contact time between each jump]; CMJ: Countermovement Jump; DJ48: Drop Jump (drop height of 48cm); MVC: Maximal Volunteer Contraction; RFD: Rate of Force Development, defined as the rise in force over time in the early phase (0-100ms) of muscle contraction (AForce / ATime). Numbers in bold express a significant correlation (P < 0,05) between test variables and total tumbling performance and total trampette performance. Associations between TG performance and mechanical muscle function Female TG expressed moderate-to-strong relationships between trampette performance and sprint capacity evaluated at 5m (r2 = 0.41, p = 0.02), 10m (r2 = 0.45, p = 0.01), 20m (r2 = 0.46, p = 0.01) and 25m (r2 = 0.45, p = 0.01). Also, 5m-sprint split-time was correlated to females' tumbling performance (r2 = 0.39, p = 0.03). In addition, a moderate-to-strong relationship was found between trampette performance and relative RFD30ms in male TG athletes (r2 = 0.44, p = 0.03). DISCUSION The aim of this study was to examining mechanical lower limb muscle function and its associations to TG performance in Danish elite male and female team gymnasts and we hypothesised that several of the test outcome would be related to TG performance. Moderate-to-strong associations were observed between sprint capacity or mechanical lower limb muscle function versus acrobatic performance in the present group of male and female elite TG athletes. Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 Anthropometrical characteristics When comparing height and body mass to that of highly trained individual gymnasts, the present male and female TG appears to be taller, heavier and with a higher body fat percentage compared to gender matched individual gymnasts (Jemni et al. 2000; Aleksic-Veljkovic et al. 2016; Rodrigues et al 2010; George et al. 2013). Rodrigues and colleagues (2013) have previously reported senior male trampoline gymnasts to have a higher fat percentage and greater body mass compared to artistic gymnasts. It has been argued that a small body size in individual gymnastics is beneficial for competitive performance (Bale & Goodway 1990; George et al. 2013). The notion that TG athletes were considerable taller and heavier than reported for individual gymnasts may be explained by differences in equipment and specific performance disciplines between TG and individual gymnastics. Thus, the rebounding equipment in TG may allow the gymnast to use a relatively longer contact time than possible in individual gymnastics, which might be beneficial for vertical impulse generation when the body mass is greater than typically seen in artistic gymnasts. Sprint capacity TG gymnasts have a run-up distance of maximum 25m in order to accelerate to the horizontal approach speed needed for performing their acrobatic drills. The present group of TG gymnasts showed faster sprint split-times compared to athletes in intermittent sports like i.e. football, hockey, netball who also rely on high sprinting velocity (Tanner & Gore 2013), indicating that high anaerobic muscle power is important in TG. In TG, the surface of the trampette is normally set to an angle of 20-30° relative to horizontal, which makes it possible for the TG gymnast to effectively convert a high horizontal run-up speed into a high vertical speed in the take-off phase, in turn resulting in a long flight time and high vertical jump height. In theory, an increased run-up speed will lead to a greater vertical jump height, due to a gain in vertical take-off velocity. However, in real-life settings even highly experienced TG gymnast may not benefit entirely from maximal running velocity during the runup. The gymnast must overcome the reactive force during the time of contact with the trampette prior to take-off. If the achieved approach velocity is too high, the gymnast may find it difficult to reach the preferred hip and knee joint angle position at the instant of eccentric-to-concentric SSC transition as well as at the instant of take-off, thereby not reaching optimal vertical height of the jump. Thus, it is unlikely for TG athletes to use their maximal sprint capacity. In contrast, in competitive team sports such as football and team handball the ability to repeat maximal sprints during a match generally is considered a highly important factor (Bangsbo et al. 2006). Sprint capacity at 5, 10, 20 and 25m were moderate-to-strongly associated with trampette performance in female TG athletes, explaining 41-46% of the observed variance in TG scores. Further, a correlation between 5m sprint capacity and tumbling performance was observed in female TG athletes indicating that high acceleration capacity is important for executing drills with high difficulty in acrobatic performance, at least among females. Thus, gymnasts with a strong acceleration capacity may be able to obtain more optimal control of the magnitude and direction of the impact force vector exerted on the trampette throughout the take-off phase. Vertical Jumping Capacity Trampette jumping and tumbling drills are fundamental aspects in TG-training. To evaluate different aspects of jumping performance, a range of vertical jump tests were used in the present study such as CMJ, drop jumping (DJ) and reactive jumping (RJ). During CMJ testing Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 duration of the take-off is often longer than observed when performing rapid DJ from a given drop height (Bobbert et al.1986) and longer than in RJ where the athletes intend to achieve a short take-off time (<250ms) combined with a high vertical jump height. These different jump types may be used to indicate the ability of the gymnast to produce SSC muscle power. CMJ testing represents the simplest way to assess anaerobic SSC leg muscle power, while DJ testing also reflects eccentric power generation that transits into a reboundjump movement while also monitoring the concurrent concentric power production for the lower limb extensors. RJ reflects the ability to repeat vertical jumping of maximal flight height using a short takeoff time, which may be relevant for tumbling performance. During the present application of CMJ, male TG athletes demonstrated relative high jumping capacity (0.51m BCM flight height) accompanied by substantial amounts of maximal muscle power generation in the concentric take-off phase (59W/kg). Our female TG athletes showed corresponding values of 0.41m and 48W/kg, which seem substantially higher than previously reported by Donti and colleagues in female artistic gymnasts (0.30m and 40W/kg, respectively) (Donti et al. 2014). Also, in comparison to Danish male trampoline gymnasts, the present male TG athletes demonstrated greater vertical jump height during CMJ (Jensen et al. 2013). Further, when compared to a group of mixed athletes (ski jumpers, alpine and freestyle skiers, snowboarders, and gymnasts) (Hilfiker et al. 2007), the present male and female TG athletes were within the same range of maximal vertical jump height (~0.32-0.47m) and leg extensor peak power (~45-65W/kg). This comparison indicates that TG athletes have relatively strong SSC jumping (CMJ) capabilities, which might be due to large amounts of TG training with focus on ballistic-type SSC movements. The DJ48 test was chosen to simulate the impact and take-off phase in the trampette and the tumbling drills. Male and female TG athletes demonstrated reduced DJ rebound height (but not peak power production) compared with CMJ, which may not be surprising due to the high eccentric impact loads imposed on the leg extensor muscles during the deceleration and acceleration phases of the DJ (Young 1995; Tanner & Gore 2013). In the present study males and females jumped from the same drop height (0.48 m). Ideally, DJ height should have been individually adjusted in order to assess and compare the capacity for absorbing kinetic energy and impulse during the rebound phase to increase rebound-jumping height across different participants (Young 1995; Tanner & Gore 2013; Taube et al. 2012). Suchomel and colleagues (2016) previously investigated the DJ abilities of male elite junior artistic gymnasts (~15 years). Compared to these junior gymnasts (rebound-jump height ~0.30m), our male TG athletes showed a markedly higher rebound-jump height (~0.43m). When compared to elite male volleyball, basketball, soccer players and recreational athletes (drop jump rebound height ~0.32m) (Cappa & Behm 2013), male TG athletes also appeared to jump higher. The RJ test was implemented to study the ability of rapidly transforming a landing to a take-off in a series of repetitive jumps, simulating performance in the tumbling discipline. It was hypothesised that team gymnasts would be comparable to track and field jumpers, as jumping exercise of various forms comprise a large part in the training performed in TG. The reactive strength index (RSI) was calculated as the ratio between average jump height and take-off time (Young 1995). Surprisingly, compared to athletics, the present gymnasts demonstrated a 58% lower RSI compared to track and field jumpers, indicating that TG athletes are not equally explosive or powerful in the transmission Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 between landing and take-off compared to state-of-the-art athletes in this field. It is possible that the relatively soft equipment used in tumbling causing the gymnasts to get used to employing a longer contact time (>250ms) between successive landings performed in an acrobatic combination, compared to the contact times observed in track and field sprinting and high jumping (<250ms). Maximal lower limb muscle strength and RFD High maximal muscle force capacity would seem a desirable feature for elite TG gymnasts, due to the high take-off forces generated during the impact with the trampette and tumbling track prior to takeoff, and the impact forces during the subsequent landing phase (McNitt 1993). The present male TG athletes demonstrated MVC values that were comparable to that observed in newly drafted males prior to initiation of military training (age 20 years, height 175cm, force 2884N) (Asmussen & Heeb0l-Nielsen 1961). The present group of female TG athletes had MVC values that were slightly higher than previously reported in female students (age 20 years, height 165cm, force 2099N) (Asmussen & Heeb0l-Nielsen 1961). MVC-normalized RFD (%MVC/s) represents a measure of the athlete's ability to express the maximal force-generating capacity in an explosive movement (Maffiuelti et al. 2016). This ability appears to be enhanced in track and field power athletes (determined at 0-50ms from onset of muscle contraction) compared to non-athlete (Tillin et al. 2010). This difference could be an indication of qualitative adaptations in power athletes, who may be characterized by higher maximal motor unit discharge rates and elevated proportions of type II muscles fibres (Maffiuelti et al. 2016). These possible adaptations and inherent neuromuscular properties have been suggested to have important influence on athletic performance (Tillin et al. 2010). In the present study, male TG gymnasts with superior performance in the trampette were also able to produce a relatively higher proportion of their maximal force capacity when evaluated in the most initial phase of muscle contraction (0-30ms). This early-phase RFD-component is likely important in the initial phase of trampette contact, due to the rapid extension of the knees and hips at the instant of contact. This rapid extension is performed to produce as much force as possible to produce a maximal reactive breaking impulse, to enable a maximal amount of elastic energy storage in the trampette for use (recoil) in the subsequent take-off phase. In summary, female elite TeamGym athletes demonstrated moderate-to-strong associations between TG performance and sprint capacity while also positively related to relative rate of force development in male athletes. In addition, TG athletes showed high sprint capacity while also characterized by high muscle power production and performance outcome during selected jumping tasks indicating that high running velocity and superior power production may be important in elite TeamGym. However, as only moderate-to-strong associations were found, performance in elite TeamGym must also rely on factors other than isolated mechanical muscle function. Limitations The chosen test battery might not fully complete to detect which physiological factors might best describe the sport in team gymnasts. Sport specificity could be improved by using a more specific foot position during the unilateral strength test as most gymnastic skills are performed with both feet close together at take-off, and therefore the unilateral leg press test might not be the ideal test for performance in TG. Also, although the gymnasts were high skilled, the results might be improved by including familiarization trials. Another limitation would be, that only tests Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 involving lower body were included. Kaldas and colleagues (2017) showed a significant correlation between performance level in artistic gymnastics and a push-up test (r2 = 0.91) as well as a pull-up test (r2 = 0.80). In future studies of TeamGym, upper body strength is recommended to be included in the battery. As in studies with elite athletes, the number of participants is limited to the very best. However, the strength is, that these athletes are supposed to be representative for top best sport performance. CONCLUSIONS This study shows that team gymnasts should focus to reach a high vertical jumping capacity, accompanied by strong sprint and acceleration skills in order to improve the physiological and biomechanical basis for their TG performance. Accordingly, it is recommended that TG athletes become exposed to plyometric jump training as well as heavy resistance training to form a strong base of muscle strength and power capacity as a prerequisite for developing the specific acrobatic expertise involved in TeamGym performance. ACKNOWLEDGEMENT Authors would like to thank the volunteers who participated in the study. REFERENCES Aagaard, P., Simonsen, E.B., Andersen, J.L., Magnusson, P., Dyhre-Poulsen, P. (2002). Increased rate of force development and neural drive of human skeletal muscle following resistance training. J Appl Physiol 93(4),1318-1326. Aleksic-Veljkovic, A., Velickovic, S., Sands, W.A., Burovic, D., Trajkovic, N. (2016). Explosive power in gymnasts: Is there any scientific basis for gender differences? Facta Universitatis: Series Physical Education & Sport 14(1),121-126. Asmussen, E., Heeb0l-Nielsen, K. (1961). Isometric muscle strength of adult men and women. Communications, 43-43. Baca, A. (1993). A comparison of methods for analyzing drop jump performance. / Comparaison de differentes methodes d'analyse de la performance lors de sauts en profondeur. Med Sci Sports Exerc 31(3),437-442. Bale, P., Goodway, J. (1990). Performance variables associated with the competitive gymnast. / Variables de performance du gymnaste de competition. Sports Medicine 10(3),139-145. Bangsbo, J., Mohr, M., Krustrup, P. (2006). Physical and metabolic demands of training and match-play in the elite football player. J Sports Sci 24(7),665-674. Bobbert, M.F., Mackay, M., Schinkelshoek, D., Huijing, P.A., van Ingen Schenau, G.J. (1986). Biomechanical analysis of drop and countermovement jumps. Eur J Appl Physiol Occup Physiol 54(6):566-573 Cappa, D.F., Behm, D.G. (2013). Neuromuscular Characteristics of Drop and Hurdle Jumps with Different Types of Landings. J Strength Cond Res 27(11), 3011-3020. Caserotti, P., Aagaard, P., Buttrup Larsen, J., Puggaard, L. (2008). Explosive heavy-resistance training in old and very old adults: changes in rapid muscle force, strength and power. Scand J Med Sci Sports 18(6),773-782. Donti, O., Tsolakis, C., Bogdanis, G.C. (2014). Effects of Baseline Levels of Flexibility and Vertical Jump Ability on Performance Following Different Volumes of Static Stretching and Potentiating Exercises in Elite Gymnasts. J Sports Sci Med 13(1), 105-113. Elbaek, L., Froberg, K. (1993). Specific physical training paramaters in relation to danish team gymnastics (Les parametres d'entrainement physique specifique rapportes a l'equipe danoise de Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 gymnastique). In. Allemagne: Koln: Sport und Buch Strauss. George, D., Elias, Z., George, P. (2013). Physiological profile of elite Greek gymnasts. Journal of Physical Education & Sport 13(1), 27-32. Harringe, M., Lindblad, S., Werner, S. (2004). Do team gymnasts compete in spite of symptoms from an injury? Br J Sports Med 38(4), 398-401. Harringe, M., Renström, P., Werner, S. (2007). Injury incidence, mechanism and diagnosis in top-level teamgym: a prospective study conducted over one season. Scand J Med Sci Sports 17(2),115-119. Hebert-Losier, K., Eriksson, A. (2014). Leg stiffness measures depend on computational method. J Biomech 47(1),115-121. Hilfiker, R., Hübner, K., Lorenz, T., Marti, B. (2007). Effect of drop jumps added to the warm-up of elite sport athletes with high capacity for explosive force development. J Strength Cond Res 21(2), 550-555. Jemni, M., Friemel, F., Lechevalier, J.M., Origas, M. (2000). Heart rate and blood lactate concentration analysis during a high-level men's gymnastics competition. / Analyse de la frequence cardiaque et du lactate sanguin lors d ' une competition de haut niveau en gymnastique masculine. J Strength Cond Res 14(4), 389-394. Jensen, P., Scott, S., Krustrup, P., Mohr, M. (2013). Physiological responses and performance in a simulated trampoline gymnastics competition in elite male gymnasts. J Sports Sci 31(16), 1761-1769. Kaldas, J., Bisson, C., Hogue, A-C., Apinis, C., Berbiche, D., Gaudrealt, N. (2017). Construct validity and inter-rater reliability of the Gymnastic Functional Measurement Tool in the classification of female competitive gymnasts in Canada. Physical Therapy in Sport 28, 9-14. Lund, S.S., Myklebust, G. (2011). High injury incidence in TeamGym competition: a prospective cohort study. Scand J Med Sci Sports 21(6), 439-444. Maffiuletti, N., Aagaard, P., Blazevich, A., Folland, J., Tillin, N., Duchateau, J. (2016). Rate of force development: physiological and methodological considerations. Eur J Appl Physiol 116(6), 1091-1116 McNitt, G-J. (1993). Kinetics of the lower extremities during drop landings from three heights (Cinetique des membres inférieurs lors de la reception de sauts en profondeur a partir de trois hauteurs.). J Biomechs 26(9), 1037-1046. Rodríguez, L.A.G-L., Santana, M.V., Bedoya, J.L. (2010). Somatotipo y composición corporal en gimnastas de Trampolín masculino español de alto nivel. / Somatotype and body composition in elite male Spanish Trampoline. RICYDE Revista Internacional de Ciencias del Deporte 6(19), 141-153. Sjostrand, P., Lemmetty, H., Hughes, K., Gryga, P., Dvoracek, R., Jónsdóttir, S. (2015). European Championship in TeamGym - Seniors and Juniors Code of Points. Suchomel, T.J., Sands, W.A., McNeal, J.R. (2016). Comparison of static, countermovement, and drop jumps of the upper and lower extremities in U.S. junior national team male gymnasts. Science of Gymnastics Journal 5(1), 15-30. Tanner, R.K., Gore, C.J. (2013). Physiological Tests for Elite Athletes. 2.ed. Australia: Human Kinetics. Taube, W., Leukel, C., Lauber, B., Gollhofer, A. (2012). The drop height determines neuromuscular adaptations and changes in jump performance in stretch-shortening cycle training. Scand J Med Sci Sports 22(5), 671-683. Thorlund, J.B., Michalsik, L.B., Madsen, K., Aagaard, P. (2008). Acute fatigue-induced changes in muscle mechanical properties and neuromuscular activity in elite handball players following a handball match. Scand J Med Sci Sports 15(4), 462-472. Tillin, N.A., Jimenez-Reyes, P., Pain, M.T.G., Folland, J.P. (2010). Neuromuscular Performance of Explosive Science of Gymnastics Journal 163 Science of Gymnastics Journal Hansen O.H., Hvid L.G., Aagaard P., Jensen K.: MECHANICAL LOWER LIMB MUSCLE. Vol. 11 Issue 2: 163 - 174 Power Athletes versus Untrained Individuals. Med Sci Sports Exerc 42(4), 781-790. Young, W. (1995). Laboratory strength assessment of athletes. / Evaluation en laboratoire de la force des athletes. New Studies in Athletics 10(1), 89-96. Corresponding author: Kurt Jensen Department of Sports Science and Clinical Biomechanics University of Southern Denmark, Campusvej 55, 5220, Odense M, Denmark Phone +45 65506089 E-mail: kjensen@health.sdu.dk Science of Gymnastics Journal 163 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 IMPACT OF GYMNASTICS TRAINING ON THE HEALTH-RELATED PHYSICAL FITNESS OF YOUNG FEMALE AND MALE ARTISTIC GYMNASTS Iliya Kiuchukov1, Iliya Yanev1, Lubomir Petrov1, Stefan Kolimechkov12, Albena Alexandrova1, Dilyana Zaykova1, Emil Stoimenov1 1 National Sports Academy, Sofia, Bulgaria 2 Elite Gymnastics Academy CIC - Coaching Department, London, Great Britain _Original article Abstract Artistic gymnastics can be practised from an early age and develops the main components of physical fitness. The aim of this study was to assess the physical fitness of young competitive artistic gymnasts from Bulgaria. A total of 161 gymnasts (81 females and 80 males), who were divided into three groups (from 5-8, 9-11, and 12-15 years of age), with sports experience from 12 to 180 months, took part in this study. All of the participants completed the extended version of the Alpha-Fit physical fitness test battery, with European norms being applied to calculate percentile scores for each fitness test. The height-for-age percentile scores in the groups between the ages of 9-11 and 12-15 were significantly lower from the 50th percentile of the international norms, both for male and female gymnasts. Gymnasts showed substantially lower body fat, and only one gymnast was assessed as overweight, with two being classified as obese. The percentile scores of the standing long jump and the 4x10 m SRT in the groups were significantly greater than the 50th percentile of the available European norms. The percentile scores of the VO2max in all female groups were also higher than the 50th percentile of the European norms, while those for males did not differ from the 50th percentile, except in the 5-8 age range. Artistic gymnastics improves the physical fitness components and positively influences children's physical development. Both female and male artistic gymnasts had better physical fitness in most parameters, in comparison with their peers. Keywords: physical fitness, artistic gymnasts, gymnastics, alpha-fit. INTRODUCTION Artistic gymnastics is one of the few sports which can be practised from a very young age, and which develops different components of physical fitness. Pupils learn basic elements which are of great significance, especially for children's orientation in space, such as jumps and leaps, hanging, rotating, crawling, rolling, etc. (Pajek, Cuk, Kovac, & Jakse, 2010). Health-related physical fitness is a major factor in children's health, and it has a multidimensional structure consisting of the following components: body composition, musculoskeletal fitness, Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 motor fitness (speed, agility and coordination), and cardiorespiratory fitness (ALPHA, 2009; Artero et al., 2011; Ruiz et al., 2009; Ruiz et al., 2010). There are different fitness test batteries, such as Alpha-fit, Eurofit, FitnessGram, etc., which are applied around the world in order to assess physical fitness in children and adolescents (Kolimechkov, 2017). Based on many longitudinal and cross-sectional studies, a wide range of authors define the Alpha-fit test battery as an ideal tool, encompassing a valid, reliable and safe set of tests for the assessment of physical fitness in children and adolescents (Cvejic, Pejovic, & Ostojic, 2013; Romero et al., 2010; Ruiz et al., 2010; Santos & Mota, 2011). Optimal health and level of physical fitness are essential for all gymnasts in order to be able to effectively and accurately perform varied elements and routines. The physical fitness assessment can provide information which allows us to track the impact of the sport on each gymnast's health (British Gymnastics, 2015). There are currently around 700 artistic gymnasts who are registered with the Bulgarian Gymnastics Federation, and more than 600 of them are under the age of 17. Therefore, the aim of this study was to assess the physical fitness levels of a representative sample of young competitive artistic gymnasts from different regions in Bulgaria, and show the impact of the training on the gymnasts' health by using the Alpha-Fit test battery. METHODS This study included a representative sample of 25% of all registered young artistic gymnasts. Thus, this study involved the participation of 161 artistic gymnasts (81 females and 80 males) who regularly took part in, or were preparing for, competitions. The gymnasts were between the ages of five and fifteen, from four different cities in Bulgaria (Sofia, Blagoevgrad, Veliko Tarnovo and Ruse) representing nine different gymnastics clubs which are registered with the Bulgarian Gymnastics Federation. The average sports experience of all participants in artistic gymnastics was 3.6 years (from 1 to 15 years). The gymnasts were divided into three age groups (from 5 to 8, 9 to 11, and 12 to 15 years of age) and according to gender. Institutional ethics approval for this research was granted by the National Sports Academy in Sofia, and informed consent was obtained from the parent/guardian of each gymnast. All artistic gymnasts completed the extended version of the Alpha-Fit physical fitness test battery (ALPHA, 2009), which included the following anthropometric measurements and tests: height, weight, waist circumference, triceps and subscapular skinfolds to assess body composition; handgrip strength and standing long jump to assess musculoskeletal fitness; the 4x10 m shuttle run test (4x10 m SRT) to assess motor fitness; and the 20 m shuttle run test (20 m SRT) to assess cardiorespiratory fitness. Stature was measured to the nearest 0.1 cm with a stadiometer, body mass was recorded to within an accuracy of 0.1 kg by using the Omron BF511 electronic scale, and waist and arm circumferences were measured to the nearest 0.1 cm with the Lufkin W606PM tape measure. All of the measurements were recorded by following the anthropometric procedures thoroughly (NHNES, 2007; Piwoz & Viteri, 1985). Body mass index (BMI) was calculated as: body mass in kilograms/stature in metres squared. In addition, the percentile scores for each gymnast's height, weight and BMI were computed and assessed by using the WHO AnthroPlus specialised software, provided by the World Health Organisation (WHO, 2011). The following classification of the BMI percentile scores (PRs) for children and adolescents between the ages of 5 and 19, provided by the WHO, was applied: Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 BMI > 85th PRs is classified as 'overweight'; BMI > 97th PRs is 'obese'; BMI < 15th PRs is 'thinness'; and BMI < 3rd PRs is 'severe thinness' (WHO, 2007a). Waist-to-height ratio (WHtR) was calculated by dividing waist circumference (cm) by height (cm), and the simple cut-off of WHtR = 0.500 was used to assess increased health risk in children relating to an excessive accumulation of body fat on the upper body (Ashwell & Hsieh, 2005; McCarthy & Ashwell, 2006). The triceps and subscapular skinfolds were measured with the Lange Skinfold Caliper, Beta Technology Inc, Cambridge to an accuracy of 1 mm. Body fat percentage (%Fat) was determined by the sum of the two skinfolds, using Slaughter's equations (Heyward & Stolarczyk, 1996; Slaughter et al., 1988), which are highly recommended for children and adolescents because of the accuracy and simplicity of this method (ALPHA, 2009; Boye et al., 2002; Laurson, Eisenmann, & Welk, 2011). Body fat percentile scores were computed by using the recent international norms for Caucasian children and adolescents (McCarthy, T.J. Cole, T. Fry, S.A. Jebb, & Prentice, 2006). In order to classify %Fat the following cut-offs were applied: %Fat > 85th PRs is classified as 'overweight'; %Fat > 95th PRs is 'obese'; and %Fat < 2nd PRs is 'underfat' (McCarthy et al., 2006). The upper arm muscle area (UAMA) was calculated from the arm circumference and the triceps skinfold by using the following formula (Addo, Himes, & Zemel, 2017): UAMA = [Arm circumference - (n x triceps skinfold)]2 + 4 x n Percentile scores for the UAMA were computed for each gymnast by using the recent norms for children and adolescents (Addo et al., 2017). In addition, the relative UAMA (cm2/kg) was also calculated by dividing the UAMA (cm2) by body mass (kg). Handgrip strength was measured for both hands by using the SH5001 Hydraulic Hand Dynamometer to assess upper body isometric strength. The elbow of the tested hand was fully extended and the testing procedure was strictly followed (ALPHA, 2009; NHANES, 2013). The relative handgrip strength for each participant was also calculated by dividing the average handgrip strength (kg) by the body mass (kg). The standing long jump test was performed to assess lower body explosive strength on a non-slippery hard surface, and the test was recorded to within an accuracy of 1 cm. The distance was measured from the take-off line to the point where the back of the heel, nearest to the take-off line, lands on the ground (ALPHA, 2009). Percentile scores for the average handgrip strength and the standing long jump were computed by using the available European norms for children (Miguel-Etayo et al., 2014) and adolescents (Ortega et al., 2011). Linear interpolations and extrapolations between the existing European norms were applied in order to compute percentile scores for those ages which were not published, these being 5, 10, 11 and 12-year-old children (Kolimechkov, Petrov, & Alexandrova, 2018). The 4x10m SRT at maximum speed was performed to measure speed of movement, agility and coordination, in accordance with the standard procedure described in the Alpha-fit test battery (ALPHA, 2009). The test was recorded in seconds by using a stopwatch to an accuracy of 0.1 sec. The percentile scores of the results were calculated by using the available European norms for children (Roriz De Oliveira, Seabra, Freitas, Eisenmann, & Maia, 2014) and adolescents (Ortega et al., 2011). Percentile scores for the missing years (5, 11 and 12-year-old children), in which there was a gap in the norms, were computed by using linear interpolations and extrapolations of the existing Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 European percentiles (S. Kolimechkov et al., 2018). The estimated maximal oxygen uptake (VO2max) was calculated by using a modified version of the 20 m SRT in order to assess the cardiorespiratory fitness of the artistic gymnasts. The modified test required running between two lines which were 10 m apart, in time with an audio signal, instead of the original 20 m. This modification made the administration of the test more convenient when conducted inside the gymnastics centres (on the gymnastics floor 12x12 m). An extended specialised version of the BeepShuttle Junior software (Kolimechkov, Petrov, Alexandrova, & Cholakov, 2018) was applied in order to administer the 10 m SRT. The software applies the original 1-minute protocol, which starts at a speed of 8.5 km/h and increases in speed by 0.5 km/h after each minute, as described by Leger et al. (Leger, Lambert, Goulet, Rowan, & Dinelle, 1984). Moreover, the software calculated the predicted VO2max by using Leger's equation (Leger, Mercier, Gadoury, & Lambert, 1988) in addition to the percentile score for each participant based on age- and gender-specific international norms (Miguel-Etayo et al., 2014; Tomkinson et al., 2016). All anthropometric measurements were taken twice, and the mean was used in the analyses, as recommended in the test manual of the Alpha-fit battery. The handgrip strength test, standing long jump test and 4x10 m SRT were performed twice, and the better score was used in the analyses, whilst the cardiorespiratory test was performed once (ALPHA, 2009). The statistical analyses were conducted with SPSS Statistics 19 software, using descriptive statistics and One-way ANOVA with the Bonferroni post hoc test. Statistically significant differences between the average values were evaluated at p < 0.05, and all data in the text are presented as mean ± SD. In addition, the percentile scores of the parameters were compared to the 50th percentile by using one sample t-test. Cohen's effect size was calculated for those parameters which differed significantly from the 50th percentile, and the following classification was applied: d (0.01) = very small, d (0.20) = small, d (0.50) = medium, d (0.80) = large, d (1.20) = very large, and d (2.00) = huge (Sawilowsky, 2009). RESULTS The anthropometric parameters, their percentile scores and their effect size vs the 50th percentiles (PRs) of the female artistic gymnasts, divided into three age groups, are presented in Table 1. The group which included gymnasts between the ages of 12 and 15 has the greatest average sports experience (7 years and 6 months). Therefore, the outcomes on physical development and physical fitness from practising artistic gymnastics should be clearly evident in this group. The average frequency of the gymnastics training ranged from 4 to 5 sessions per week. The height-for-age percentile scores in the groups between the ages of 9-11 and 12-15 were significantly lower than the 50th percentile of the international norms for children and adolescents at this age provided by the World Health Organization (WHO, 2006). The average weight of the gymnasts increased gradually with age and did not differ from the average international standards. However, it should be taken into account that the World Health Organization (WHO) does not provide weight-for-age reference data for children over 10 years of age, because this indicator cannot distinguish between height and body mass at an age when many children are experiencing the pubertal growth spurt (WHO, 2007b). The body mass index (BMI) did not show significant differences from the international percentile scores, except with the first age group (5-8 years), where the Cohen's effect size was small (d=0.41). The average waist-to-height ratio in all three Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 age groups was below the boundary of 0.500, which distinguishes children at risk as far as their health is concerned (Ashwell & Hsieh, 2005; McCarthy & Ashwell, 2006). The %Fat percentile scores in all groups were greatly lower than the 50th percentile of the international norms for children and adolescents (McCarthy et al., 2006), and the Cohen's effect size was large (d=1.05) for 5-8-year-old gymnasts, very large (d=1.37) for 9-11-year-old gymnasts, and huge (d=3.17) for 12-15-year-old gymnasts, in accordance with the benchmarks provided by Cohen and Sawilowsky (Lakens, 2013; Sawilowsky, 2009). The upper arm muscle area (UAMA) percentile scores for all three groups are lower than the 50th percentile, but this difference was significant only in the group with gymnasts between the ages of 9 and 11 with a medium effect size (d=0.50). Table 1 Anthropometric parameters, their percentile scores and effect size vs the 50th percentile (PRs) of the female artistic gymnasts divided into three age groups (mean ± SD). 5-8 years (n=28) 9-11 years (n=39) 12-15 years (n=14) Age (years) 7.45 ± 0.92 10.25 ± 0.95 13.52 ± 1.28 Sports experience (months) 23.71 ± 12.91 C 36.76 ± 23.07 C 79.57 ± 38.77 Sessions per week 4.05 ± 0.88 Bc 4.86 ± 0.80 4.94 ± 0.85 Height (cm) 123.96 ± 8.26 BC 138.38 ± 8.98 C 153.13 ± 6.57 Height - percentile score 51.53 ± 31.56 39.98 ± 29.35 32.96 ± 25.07 Effect size vs 50th PRs NS 0.01 a 0.68 a Weight (kg) 25.73 ± 5.87 BC 33.22 ± 6.25 C 44.50 ± 7.55 Weight - percentile score 58.18 ± 29.20 B 50.54 ± 23.33** -* Effect size vs 50th PRs NS NS BMI (kg/cm2) 16.54 ± 1.99 c 17.23 ± 1.86 c 18.84 ± 1.90 BMI - percentile score 61.09 ± 27.29 53.93 ± 25.70 44.81 ± 22.96 Effect size vs 50th PRs 0.41 a NS NS Arm circumference (cm) 18.31 ± 2.05 BC 20.02 ± 1.70 C 22.16 ± 1.59 Waist circumference (cm) 53.66 ± 5.87 bC 57.99 ± 3.79 c 61.84 ± 4.45 Waist-to-height ratio 0.43 ± 0.04 c 0.42 ± 0.03 0.40 ± 0.02 Subscapular skinfold (mm) 6.95 ± 3.00 7.13 ± 2.61 7.88 ± 1.67 Triceps skinfold 10.52 ± 3.72 11.38 ± 4.21 9.61 ± 3.29 % Fat 16.27 ± 4.60 17.14 ± 4.73 16.54 ± 3.88 % Fat - percentile score 21.04 ± 27.53 17.32 ± 23.77 10.39 ± 12.51 Effect size vs 50th PRs 1.05 A 1.37 A 3.17 A UAMA (cm2) 18.07 ± 3.42 BC 21.67 ± 3.77 C 29.34 ± 4.64 UAMA - percentile score 43.20 ± 26.64 38.29 ± 23.25 39.61 ± 23.94 Effect size vs 50th PRs NS 0.50 a NS Relative UAMA (cm2/kg) 0.71 ± 0.10 0.66 ± 0.11 0.67 ± 0.12 * WHO does not provide weight-for-age reference data for children older than 10 years of age (WHO, 2007b). ** n=19 because 20 out of 39 female gymnasts were older than 10 (see *). a - p < 0.05 vs 50th PRs; a - p < 0.01 vs 50th PRs; A - p < 0.001 vs 50th PRs; b - p < 0.01 vs 9-11 years; B - p < 0.001 vs 9-11 years; c - p < 0.05 vs 12-15 years; c - p < 0.01 vs 12-15 years; C - p < 0.001 vs 12-15 years; NS - not significant Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 Table 2 Results of the handgrip strength test, standing long jump, 4x10 m SRT, 20m SRT, their percentile scores and effect size (vs 50th percentile) of all female artistic gymnasts (mean ± SD)._ 5-8 years (n=28) 9-11 years (n=39) 12-15 years (n=14) Musculoskeletal Fitness: Upper body strength Handgrip strength test* (kg) 8.87 ± 2.91 BC 13.72 ± 4.24 C 20.05 ± 4.18 Handgrip strength test (percentile score) 32.32 ± 30.27 31.65 ± 28.57 26.88 ± 21.09 Effect size vs 50th PRs 0.58 a 0.64 A 1.10 a Relative handgrip strength 0.34 ± 0.08 bc 0.41 ± 0.11 0.45 ± 0.06 (kg/ kg body weight) Musculoskeletal Fitness: Lower body strength Standing long jump (cm) 129.46 ± 17.95 BC 160.83 ± 20.92 C 195.71 ± 15.29 Standing long jump 83.45 ± 20.27 88.65 ± 16.06 96.26 ± 4.65 (percentile score) Effect size vs 50th PRs 1.65 A 2.41 A 9.95 A Motor Fitness 4x10 m shuttle run test (sec) 13.81 ± 0.94 BC 12.76 ± 1.12 C 11.57 ± 0.64 4x10 m shuttle run test (percentile score) 74.38 ± 17.44 69.26 ± 25.45 83.30 ± 14.05 Effect size vs 50th PRs 1.40 A 0.76 A 2.37 A Cardiorespiratory Fitness VO2max (ml/kg/min) 49.06 ± 1.99 bC 46.48 ± 2.92 C 43.25 ± 3.49 VO2max (percentile score) 79.23 ± 14.90 B 61.19 ± 22.35 72.42 ± 16.48 Effect size vs 50th PRs 1.96 A 0.50 a 1.36 A * - values expressed as average of right and left hands a - p < 0.01 vs 50th PRs; A - p < 0.001 vs 50 th PRs; b - p < 0.05 vs 9-11 years; b - p < 0.01 vs 9-11 years; B - p < 0.001 vs 9-11 years; c - p < 0.01 vs 12-15 years; C - p < 0.001 vs 12-15 years; NS - not significant Table 3 Anthropometric parameters, their percentile scores and effect size vs the 50th percentile (PRs) of the male artistic gymnasts divided into three age groups (mean ± SD). 5-8 years (n=35) 9-11 years (n=28) 12-15 years (n=17) Age (years) 7.45 ± 1.09 10.29 ± 0.93 13.48 ± 1.10 Sports experience (months) 27.66 ± 14.60 BC 49.57 ± 16.70 C 84.94 ± 31.23 Sessions per week 4.51 ± 0.68 bc 5.06 ± 0.62 4.98 ± 0.61 Height (cm) 122.64 ± 8.28 BC 135.11 ± 7.60 C 148.06 ± 9.67 Height - percentile score 41.71 ± 27.49 C 31.36 ± 23.04 c 11.99 ± 9.54 Effect size vs 50th PRs NS 0.81 A 3.10 A Weight (kg) 23.98 ± 3.65 BC 30.73 ± 4.95 C 38.66 ± 8.71 Weight - percentile score 47.43 ± 25.90 B 39.22 ± 27.61** -* Effect size vs 50th PRs NS NS BMI (kg/cm2) 15.88 ± 1.31 c 16.74 ± 1.65 17.38 ± 1.76 BMI - percentile score 53.42 ± 25.35 c 48.76 ± 26.52 c 28.64 ± 22.16 Effect size vs 50th PRs NS NS 0.96 a Arm circumference (cm) 17.87 ± 1.58 BC 20.23 ± 1.80 c 21.88 ± 2.13 Waist circumference (cm) 54.79 ± 3.32 bC 57.87 ± 4.83 c 62.14 ± 4.43 Waist-to-height ratio 0.45 ± 0.03 bC 0.43 ± 0.03 0.42 ± 0.02 Subscapular skinfold (mm) 5.49 ± 1.54 5.99 ± 1.70 5.97 ± 1.06 Triceps skinfold 8.15 ± 2.25 8.45 ± 2.99 c 6.43 ± 2.20 % Fat 13.22 ± 3.20 c 13.13 ± 3.76 c 10.31 ± 3.00 % Fat - percentile score 15.56 ± 23.78 15.34 ± 23.46 5.93 ± 9.12 Effect size vs 50th PRs 1.45 A 1.48 A 4.83 A UAMA (cm2) 18.81 ± 3.48 BC 24.87 ± 5.29 C 31.80 ± 7.65 UAMA - percentile score 36.52 ± 27.82 51.67 ± 30.42 31.86 ± 22.81 Effect size vs 50th PRs 0.48 a NS 0.80 a Relative UAMA (cm2/kg) 0.79 ± 0.12 0.82 ± 0.18 0.83 ± 0.13 * WHO does not provide weight-for-age reference data for children older than 10 years of age (WHO, 2007b). ** n=14 because 14 out of 28 male gymnasts were older than age of 10 (see *). a - p < 0.01 vs 50th PRs; A - p < 0.001 vs 50th PRs; Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 b - p < 0.05 vs 9-11 years; b - p < 0.01 vs 9-11 years; B - p < 0.001 vs 9-11 years; c - p < 0.05 vs 12-15 years; c - p < 0.01 vs 12-15 years; C - p < 0.001 vs 12-15 years; NS - not significant Table 4 Results of the handgrip strength test, standing long jump, 4x10 m SRT, 20m SRT, their c^^vnc zra/7 r>-ffr>^1 DiV/. /no <\Oth ns>vs>sm+ils>\ n-fsvll m^yIn rn-HcH/^ mmui/Trio _1_ ) 5-8 years (n=35) 9-11 years (n=28) 12-15 years(n=17) Musculoskeletal Fitness: Upper body strength Handgrip strength test* (kg) 9.69 ± 2.66 BC 15.71 ± 4.18 C 25.18 ± 5.89 Handgrip strength test (percentile score) 31.90 ± 24.47 37.29 ± 27.00 36.72 ± 27.77 Effect size vs 50th PRs 0.74 A 0.47 a NS Relative handgrip strength (kg/ kg body weight) 0.40 ± 0.09 bC 0.51 ± 0.12 C 0.66 ± 0.15 Musculoskeletal Fitness: Lower body strength Standing long jump (cm) 140.92 ± 21.18 BC 174.75 ± 21.12 C 208.03 ± 19.08 Standing long jump (percentile score) 87.01 ± 13.03 89.87 ± 14.25 91.82 ± 5.29 Effect size vs 50th PRs 2.84 A 2.80 A 7.91 A Motor Fitness 4x10 m shuttle run test (sec) 13.99 ± 1.44 BC 12.22 ± 1.08 11.32 ± 0.72 4x10 m shuttle run test (percentile score) 54.09 ± 22.89 65.66 ± 26.76 71.35 ± 22.45 Effect size vs 50th PRs NS 0.59 a 0.95 a Cardiorespiratory Fitness VO2max (ml/kg/min) 49.13 ± 3.12 c 48.38 ± 4.20 46.14 ± 3.82 VO2max (percentile score) 65.69 ± 21.20 59.23 ± 26.86 56.93 ± 20.30 Effect size vs 50th PRs 0.74 A NS NS * - values expressed as average of right and left hands a - p < 0.05 vs 50th PRs; a - p < 0.01 vs 50th PRs; A - p < 0.001 vs 50th PRs; b - p < 0.01 vs 9-11 years; B - p < 0.001 vs 9-11 years; c - p < 0.05 vs 12-15 years; C - p < 0.001 vs 12-15 years; NS - not significant The results of the handgrip strength test, standing long jump, 4x10 m SRT, 20m SRT, their percentile scores and their effect size (vs the 50th percentile) of all female artistic gymnasts are presented in Table 2. The handgrip strength percentile scores in all groups were lower than the 50th percentile of the international norms for children and adolescents, and the Cohen's effect size was medium for 5-8-year-old gymnasts (d=0.58) and 9-11-year-old gymnasts (d=0.64), and large (d=1.10) for 12-15-year-old gymnasts. The standing long jump percentile scores in all three groups were significantly higher than the 50th percentile of the available European norms for children and adolescents at this age (Miguel-Etayo et al., 2014; Ortega et al., 2011), and the effect size was very large (d=1.65) for those female gymnasts who were from 5 to 8 years of age, and huge for the older ones (d=2.41 for the 911-year-old gymnasts and d=9.95 for the 12-15-year-old gymnasts). The 4x10 m SRT percentile scores were also significantly higher than the 50th percentile in all groups, and the effects sizes were: very large (d=1.40 for the 5-8-year-old gymnasts), medium (d=0.76 for the 9-11-year-old gymnasts), and huge (d=2.37 for the 12-15-year-old gymnasts). The percentile scores of the VO2max obtained by the modified 20 m SRT in all three groups were also higher than the 50th percentile of the European norms, and the effect size was very large (d=1.96 for the 5-8-year-old gymnasts, and d=1.36 for the 12-15-year-old gymnasts), and medium (d=0.50 for the 9-11-year-old gymnasts). The anthropometric parameters, their percentile scores and their effect size vs the 50th percentiles (PRs) of the male artistic gymnasts, divided into three age groups, are presented in Table 3. As expected, the group which included the oldest male gymnasts had the greatest average sports experience (7 years), which was also registered with the female gymnasts. The Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 average frequency of the gymnastics training ranged between 4 and 6 sessions per week. The height-for-age percentile scores in the groups between the ages of 911 and 12-15 were significantly lower than the 50th percentile of the international norms, with a large (d=0.81) and huge effect size (d=3.10), respectively. The average weight of the male gymnasts did not differ significantly from the international standards. The body mass index (BMI) did not show significant differences from the international percentile scores, except in the third group (12-15 years), where the effect size was large (d=0.96). The average waist-to-height ratio in all three age groups was within healthy norms, as was registered with the female gymnasts. The %Fat percentile scores in all male groups were substantially lower than the 50th percentile of the international norms for children and adolescents, and the effect size was very large both for the 5-8-year-old gymnasts (d=1.45) and the 9-11-year-old gymnasts (d=1.48), and huge for 12-15-year-old gymnasts (d=4.83). UAMA percentile scores for two of the groups (5-8-year-old and 12-15-year-old male gymnasts) were significantly lower than the 50th percentile, with a small (d=0.48) and large effect size (d=0.80), respectively. The results of the handgrip strength test, standing long jump, 4x10 m SRT, 20m SRT, their percentile scores and their effect size (vs the 50th percentile) of all male artistic gymnasts are presented in Table 4. The handgrip strength percentile scores in all groups were lower than the 50th percentile of the international norms for children and adolescents, but differ significantly only in the groups with the younger gymnasts: 5-8 years of age and 911 years of age. The effect size was medium (d=0.74) and small (d=0.47), respectively. The standing long jump percentile scores in all three groups were significantly greater than the 50th percentile of the available European norms, and the effect size was huge for all three groups (d=2.84 for 5-8-year-old gymnasts, d=2.80 for the 9-11-year-old gymnasts and d=7.91 for the 12-15-year-old gymnasts). The 4x10 m SRT percentile scores were significantly higher than the 50th percentile in two of the groups: 9-11-year-old male gymnasts with a medium effect size (d=0.59), and 12-15-year-old male gymnasts with a large effect size (d=0.95). The percentile scores of the VO2max did not differ from the 50th percentile of the European norms, except in the 5-8-year-old male gymnasts, where the PRs scores were significantly higher, and the effect size was medium (d=0.74). DISCUSION The progressive decrease of the percentile scores in the height of the gymnasts (Table 1 & Table 3) is probably because of the fact that those of shorter stature are more likely to have an advantage when performing many of the gymnastics exercises. For instance, top level male artistic gymnasts (n=10) from the Swiss National team had an average stature of 168.6 ± 4.5 cm (Hubner & Scharer, 2015), which is below the average height (178.2 cm), in accordance with national norms for Swiss male adults (Grasgruber, Sebera, Hrazdira, Cacek, & Kalina, 2016). Cuk et al. concluded that there was no difference in the height of top level male artistic gymnasts in 1933 and those in 2000, with their average height being 168 cm (Cuk et al., 2007). Similarly, there was no significant difference in the average height of top level male artistic gymnasts in 2000 from those in 2015 (Sibanc, Kalichova, Hedbavny, Cuk, & Pajek, 2017). This lower than average stature is also seen in other studies (Benardot, 2014), where the average values of the height of young gymnasts are similar to those in our study. For instance, the height-for-age in female junior elite gymnasts progressively dropped from the 48th to the 20th percentile as age increased (Benardot & Czerwinski, 1991). However, Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 that does not mean that gymnastics training slows down growth. In a recent review about the role of intensive training on the growth of artistic gymnasts, Malina et al. (2013) concluded that adult height or near adult height of artistic gymnasts of both genders is not compromised by intensive gymnastics training at a young age or during the pubertal growth spurt. Artistic gymnasts are shorter and lighter than average, but gymnastics training does not attenuate growth of upper or lower body segments (Malina et al., 2013). In fact, gymnastics is a unique sport that provides competitive opportunities for the smallest and lightest athletes in a world where many sports are clearly biased in favour of athletes who are tall and/or big (Sands, 1999). Although the BMI is the most popular method and is widely used for the assessment of body composition (Flegal, Tabak, & Ogden, 2006; Keys, Fidanza, Karvonen, Kimura, & Taylor, 2014; Pekar, 2011), it did not provide correct individual assessment of some of the gymnasts involved in our study, because the BMI does not distinguish between fat and muscle mass. That is why some authors highlight that the BMI is not appropriate for some groups of people, such as professional athletes, body building enthusiasts, people engaged in jobs with strenuous physical activity (Bogin & Varela-Silva, 2012) and adolescent athletes (Lutoslawska et al., 2014). Moreover, the BMI was also shown to be an inadequate indicator of weight and body composition in child athletes with greater muscle mass (Kolimechkov et al., 2013). The percentage body fat was very low, both in female and male artistic gymnasts, which is normal for children and adolescents involved in gymnastics (Jemni, 2011). The results of %Fat from our study are similar to those reviewed by Benardot, 2014, where the average %Fat for children and adolescents engaged in gymnastics ranged between 8.6% and 21.5%. The percentile scores for the handgrip strength in both female and male gymnasts were lower than the 50th percentile for their age. On the whole, artistic gymnasts have smaller body sizes (especially in older children), in comparison to those for their age and gender, than the international norms, and, therefore, the evaluation of the strength parameters will be better assessed by relative parameters. In addition, the workload in artistic gymnastics comes mainly from the gymnasts' body weight. Consequently, percentile scores of relative handgrip strength (per kg body weight), as well as percentile scores of relative UAMA (per kg body weight), will be a better way to appropriately assess both the gymnasts' muscle mass and strength. However, to the best of our knowledge, such norms for children are still not available in the literature, and we are of the opinion that these norms should be obtained in future research. The standing long jump represents the relative lower body strength and, not unexpectedly, this test witnessed better results in artistic gymnasts, because jumps in height and length are included in artistic gymnastics training. The 4x10 m SRT represents the lower body strength, speed and agility, and shows a high correlation with the standing long jump test (r = -0.73, p < 0.001 and - 0.83, p < 0.001 for girls and boys, respectively). The results of the 4x10 m SRT were also expected to be better in artistic gymnasts, because of the short distance sprints which precede gymnastic vaults and acrobatic series. Similarly, rhythmic gymnasts between the ages of 7 and 17, who completed the Alpha-fit test battery, also achieved their best results in those two tests (Montosa, Vernetta, & Lopez-Bedoya, 2018). As can be expected, the results from the standing long jump and the 4x10 m SRT in our study showed the largest effect size in the group with the most experience in gymnastics (12-15 years of age), Table 2 and Table 4. Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 Unfortunately, the Alpha-fit test battery is not fully completed in terms of available percentile scores for the evaluation of the tests results. There are still certain age groups without European percentile scores for the standing long jump test, handgrip strength test, 4x10 m SRT and 20 m SRT. Miguel-Etayo et al. also talk about this reference gap between the ages of 10 and 12 at the European level, which has to be filled in (Miguel-Etayo et al., 2014). Meanwhile, interpolated and extrapolated percentile scores of the existing data can be used in order to evaluate the results of children and adolescents at any age (S. Kolimechkov et al., 2018). The maximal oxygen uptake (VO2max) of the female artistic gymnasts decreased significantly with age, but remained significantly higher than the 50th percentile of the international age- and gender-specific norms. The effect size was very large (d=1.36) in the group with the most experience in gymnastics. The VO2max of the male artistic gymnasts also gradually decreased with age, and the values are similar to those published for artistic gymnasts in the literature, with an average VO2max of 50 ml/kg/min (Jemni, 2011). Although higher average values were reported for the American elite female gymnasts, around 60 ml/kg/min, (Noble, 1975), Jemni (2011) points out that the VO2max of international level gymnasts reported over the last 50 years remains the same, around 50 ml/kg/min (Jemni, 2011). Barantsev (1985) found out that VO2max decreases between adolescence and adulthood, with average values dropping from 53.2 ± 6.3 at age 12 to 47.2 ± 6.7 ml/kg/min at age 25 (Barantsev, 1985). Furthermore, Jemni (2011) highlights that this regression is not evident before puberty, and is associated with the higher volume and intensity of training for the strength and power required for difficult technical exercises, specifically in men's gymnastics (Jemni, 2011). The higher percentile scores for the female gymnasts are probably due to the greater average duration of the gymnastics routines (Jemni, Friemel, Le Chevalier, & Origas, 2000). In addition, on three out of the four female gymnastics apparatuses, the lower body is constantly involved, while the exercises on most of the male apparatuses are predominantly concentrated on the upper body. CONCLUSIONS Artistic gymnastics improves all health-related components of physical fitness and positively influences children's physical development. Both female and male artistic gymnasts had better physical fitness in most parameters, in comparison to the international norms for their peers. The results suggest that body fat percentage should be used instead of BMI for gymnasts in order to accurately assess their body composition. Percentile scores of relative handgrip strength and relative upper arm muscle area (UAMA) should be obtained in future research and applied in order to appropriately assess artistic gymnasts. ACKNOWLEDGEMENT This research was financially supported by Grant '08/15.02.2018' from the National Sports Academy, Sofia, Bulgaria. REFERENCES Addo, O. Y., Himes, J. H., & Zemel, B. S. (2017). Reference ranges for midupper arm circumference, upper arm muscle area, and upper arm fat area in US children and adolescents aged 1-20 y. Am J Clin Nutr, 105(1), 111-120. doi: 10.3945/ajcn.116.142190 ALPHA. (2009). The ALPHA Health-related Fitness Test battery for Children and Adolescents, Test Manual. Artero, E. G., Espana-Romero, V., Castro-Pinero, J., Ortega, F. B., Suni, J., Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 Castillo-Garzon, M. J., & Ruiz, J. R. (2011). Reliability of field-based fitness tests in youth. Int J Sports Med, 32(3), 159-169. doi: 10.1055/s-0030-1268488 Ashwell, M., & Hsieh, S. D. (2005). Six reasons why the waist-to-height ratio is a rapid and effective global indicator for health risks of obesity and how its use could simplify the international public health message on obesity. Int J Food Sci Nutr, 56(5), 303-307. doi: 10.1080/09637480500195066 Barantsev, A. (1985). Do gymnasts need to develop aerobic capacity? Soviet Sports Review, 25(1), 20-22. Benardot, D. (2014). Gymnastics. In R. Maughan (Ed.), Sports Nutrition: The Encyclopedia of Sports Medicine (Vol. VII, pp. 588-608). London: IOC Medical Commission Publication. Benardot, D., & Czerwinski, C. (1991). Selected body composition and growth measures of junior elite gymnasts. Journal of the American Dietetic Association, 91(1), 29-33. Bogin, B., & Varela-Silva, I. (2012). The Body Mass Index: the good, the bad and the horrid. Bulletin de la Societe Suisse d'Anthropologie, 18(2), 5-11. Boye, K. R., T. Dimitriou, F. Manz, E. Schoenau, C. Neu, S. Wudy, & Remer, T. (2002). Anthropometric assessment of muscularity during growth: estimating fat-free mass with 2 skinfold-thickness measurements is superior to measuring midupper arm muscle area in healthy prepubertal children. The American Journal of Clinical Nutrition, 76(3), 628632. British_Gymnastics. (2015). Level 3 Coaching Theory - Resource Pack. UK. Cuk, I., Korencic, T., Tomazo-Ravnik, T., Pecek, M., Bucar, M., & Hraski, Z. (2007). Differencies in morphologic characteristics between top level gymnasts of year 1933 and 2000. CollAntropol, 31(2), 613-619. Cvejic, D., Pejovic, T., & Ostojic, S. (2013). Assessment of physical fitness in children and adolescents. Physical Education and Sport, 11(2), 135-145. Flegal, K. M., Tabak, C. J., & Ogden, C. L. (2006). Overweight in children: definitions and interpretation. Health Educ Res, 21(6), 755-760. doi: 10.1093/her/cyl128 Grasgruber, P., Sebera, M., Hrazdira, E., Cacek, J., & Kalina, T. (2016). Major correlates of male height: A study of 105 countries. Econ Hum Biol, 21, 172-195. doi: 10.1016/j.ehb.2016.01.005 Heyward, V. H., & Stolarczyk, L. M. (1996). Applied Body Composition Assessment. USA: Human Kinetics. Hubner, K., & Scharer, C. (2015). Relationship between swallow, support scale and iron cross on rings and their specific preconditioning strengthening exercises. Science of Gymnastics Journal, 7(3), 59-68. Jemni, M. (2011). The Science of Gymnastics. London, UK: Routledge. Jemni, M., Friemel, F., Le Chevalier, G. M., & Origas, M. (2000). Heart rate and blood lactate concentration analysis during a high level men's gymnastics competition. The Journal of Strength and Conditioning Research, 14(4), 389-394. Keys, A., Fidanza, F., Karvonen, M. J., Kimura, N., & Taylor, H. L. (2014). Indices of relative weight and obesity. Int J Epidemiol, 43(3), 655-665. doi: 10.1093/ije/dyu058 Kolimechkov, S. (2017). Physical Fitness Assessment in Children and Adolescents: A Systematic Review. European Journal of Physical Education and Sport Science, 3(4), 65-78. Kolimechkov, S., Petrov, L., & Alexandrova, A. (2018). Combined sets of European physical fitness percentile scores, with appropriate interpolations, for children and adolescents for the Alpha-fit test battery. Paper presented at the 13th FIEP European Congress and 29th FIEP World Congress, Istanbul, Turkey. Kolimechkov, S., Petrov, L., Alexandrova, A., & Cholakov, K. (2018). BeepShuttle Junior: Software for the Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 Administration of the 20m Shuttle Run Test in Children and Adolescents. Journal of Advanced Sport Technology, 1(3), 3540. Kolimechkov, S., Petrov, L., Ilinova, B., Alexandrova, A., Andreeva, L., & Atanasov, P. (2013). Assessment of the physical development of pre-school and primary school children practising artistic gymnastics. Journal of Sport Science, 4, 106-115. Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs. Front Psychol, 4, 863. doi: 10.3389/fpsyg.2013.00863 Laurson, K. R., Eisenmann, J. C., & Welk, G. J. (2011). Body fat percentile curves for U.S. children and adolescents. American Journal of Preventive Medicine, 41(4 Suppl 2), S87-92. doi: 10.1016/j.amepre.2011.06.044 Leger, L., Lambert, J., Goulet, A., Rowan, C., & Dinelle, Y. (1984). [Aerobic capacity of 6 to 17-year-old Quebecois--20 meter shuttle run test with 1 minute stages]. Can JAppl Sport Sci, 9(2), 64-69. Leger, L. A., Mercier, D., Gadoury, C., & Lambert, J. (1988). The multistage 20 metre shuttle run test for aerobic fitness. J Sports Sci, 6(2), 93-101. doi: 10.1080/02640418808729800 Lutoslawska, G., M. Malara, P. Tomaszewski, K. Mazurek, A. Czajkowska, A. Keska, & Tkaczyk, J. (2014). Relationship between the percentage of body fat and surrogate indices of fatness in male and female Polish active and sedentary students. Journal of Physiological Anthropology, 33(10). doi: 10.1186/1880-6805-33-10 Malina, R. M., Baxter-Jones, A. D., Armstrong, N., Beunen, G. P., Caine, D., Daly, R. M., . . . Russell, K. (2013). Role of intensive training in the growth and maturation of artistic gymnasts. Sports Med, 43(9), 783-802. doi: 10.1007/s40279-013-0058-5 McCarthy, H. D., & Ashwell, M. (2006). A study of central fatness using waist-to-height ratios in UK children and adolescents over two decades supports the simple message--'keep your waist circumference to less than half your height'. Int J Obes (Lond), 30(6), 988-992. doi: 10.1038/sj.ijo.0803226 McCarthy, H. D., T.J. Cole, T. Fry, S.A. Jebb, & Prentice, A. M. (2006). Body fat reference curves for children. International journal of Obesity, 30, 598602. Miguel-Etayo, P., L. Gracia-Marco, F. Ortega, T. Intemann, R. Foraita, L. Lissner, . . . Moreno., L. (2014). Physical fitness reference standards in European children: the IDEFICS study. International journal of Obesity, 38, 57-66. Montosa, I., Vernetta, M., & López-Bedoya, J. (2018). Assessment of health-related fitness by the ALPHA-fitness test battery in girls and adolescents who practise rhythmic gymnastics. Journal of Human Sport and Exercise, 13(1), 1-17. NHANES. (2013). Muscle Strength Procedures Manual: National Health and Nutrition Examination Survey (NHANES). NHNES. (2007). National Health and Nutrition Examination survey (NHNES). Anthropometry procedures manual. USA: CDC. Noble, L. (1975). Heart rate and predicted Vo2 during women's competitive gymnastic routines. The Journal of sports medicine and physical fitness, 15(2), 151157. Ortega, F., E. Artero, J. Ruiz, V. España-Romero, D. Jiménez-Pavón, G. Vicente-Rodriguez, . . . Castillo, M. (2011). Physical fitness levels among European adolescents: the HELENA study. British Journal of Sports Medicine, 45, 2029. Pajek, M., Cuk, I., Kovac, M., & Jakse, B. (2010). Implementation of the gymnastics curriculum in the third cycle of basic school in Slovenia. Science of Gymnastics Journal, 2(3), 15-27. Pekar, T. (2011). Body Mass Index. IMS Magazine, Summer 2011, 21-22. Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 Piwoz, G., & Viteri, F. (1985). Food and Nutrition Bulletin. UNU, 07(4), 86. Romero, V., E. Artero, D. Jimenez-Pavon, M. Cuenca-Garcia, F. Ortega, J. Castro-Pinero, . . . Ruiz, J. (2010). Assessing Health-Related Fitness Tests in the School Setting: Reliability, Feasibility and Safety; The ALPHA Study Int J Sports Med. Roriz De Oliveira, M. S., Seabra, A., Freitas, D., Eisenmann, J. C., & Maia, J. (2014). Physical fitness percentile charts for children aged 6-10 from Portugal. The Journal of sports medicine and physical fitness, 54(6), 780-792. Ruiz, J., J. Castro-Pinero, E. Artero, F. Ortega, M. Sjostrom, J. Suni, & Castillo, M. (2009). Predictive validity of health-related fitness in youth: a systematic review. Br J Sports Med, 43(12), 909-923. Ruiz, J., J. Castro-Pinero, V. Espana-Romero, E. Artero, F. Ortega, M. Cuenca, . . . Castillo, M. (2010). Field-based fitness assessment in young people: the ALPHA health-related fitness test battery for children and adolescents. Br J Sports Med. Sands, W. (1999). Why Gymnastics? USA Gymnastics Online: Technique, 19(3). Santos, R., & Mota, J. (2011). The ALPHA health-related physical fitness test battery for children and adolescents. Nutr. Hosp., 26(6), 1199-1200. Sawilowsky, S. (2009). New Effect Size Rules of Thumb. Journal of Modern Applied Statistical Methods, 8(2), 597-599. Sibanc, K., Kalichova, M., Hedbavny, P., Cuk, I., & Pajek, M. (2017). Comparison of morphological characteristics of top level male gymnasts between the years of 2000 and 2015. Science of Gymnastics Journal, 9(2), 201211. Slaughter, M., T. Lohman, R. Boileau, C. Horswill, R. Stillman, M. Van Loan, & Bemben, D. (1988). Skinfold equations for estimation of body fatness in children and youth. Human biology, 60(5), 709-723. Tomkinson, G. R., Lang, J. J., Tremblay, M. S., Dale, M., LeBlanc, A. G., Belanger, K., & Leger, L. (2016). International normative 20 m shuttle run values from 1 142 026 children and youth representing 50 countries. Br J Sports Med. doi: 10.1136/bjsports-2016-095987 WHO (2006). WHO Child Growth Standards based on length/height, weight and age. Acta Paediatr Suppl, 450, 76-85. WHO. (2007a). BMI-for-age (5-19 years). Body Mass Index for age (5-19 years), WHO AnthroPlus software. Retrieved accessed on 20 September, 2017, from http://www.who.int/growthref/who2007 b mi_for_age/en/ WHO (2007b). Weight-for-age (5-19 years). Weight for age (5-19 years), WHO AnthroPlus software. Retrieved accessed on 20 October, 2017, from http://www.who.int/growthref/who2007_w eight for age/en/ WHO (2011). WHO Anthro for personal computers, version 3.2.2, 2011: Software for assessing growth and development of the world's children. Geneva: WHO, 2010. Corresponding author: Stefan Kolimechkov Elite Gymnastics Academy CIC - Coaching Department 2 The Broadway London N9 0TR United Kingdom of Great Britain and Northern Ireland E-mail: dr.stefan.kolimechkov@gmail.com Science of Gymnastics Journal 175 Science of Gymnastics Journal Kiuchukov I., et al.: IMPACT OF GYMNASTICS TRAINING ON THE HEALTH Vol. 11 Issue 2: 175 - 187 Science of Gymnastics Journal 175 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 HEAD-TOE DISTANCE AS A SIMPLE MEASURE TO EVALUATE AMPLITUDE OF CIRCLES ON POMMEL HORSE Toshiyuki Fujihara1, Pierre Gervais2, Gareth Irwin3 1Osaka University of Health and Sport Sciences, Osaka, Japan 2University of Alberta - Faculty of Kinesiology, Sport, and Recreation, Edmonton, Canada 3Cardiff Metropolitan University - Cardiff School of Sports, Cardiff, Great Britain _Original article Abstract To develop scientifically-valid tools to monitor performance in practice, a critical question is what to measure. On pommel horse, the importance of fundamental skills called circles is uncontroversial, and one of the key performance qualities of circles is the amplitude of the movement. Previous studies have used joint angles or the magnitude of a body part's trajectory to evaluate the amplitude, but we hypothesized that the distance between two points, namely a head and toes might be substituted despite its relative simplicity. This study examined the use of Head-Toe Distance (HTD) normalized by the gymnast's body height as a simple variable to potentially evaluate the amplitude of circles. The kinematic data of circles performed by 18 elite gymnasts were collected with a Qualisys motion capture system operating at 100 Hz. HTD and its horizontal component, HTDh, were computed along with their relationships to the outcome scores given by the official judges, as well as the other amplitude variables: the horizontal diameters of shoulder and ankle trajectories; the body flexion angle; and in the rear support position, the shoulder extension angle and the head position. The results supported HTDh, rather than HTD, for its potential usage as a single variable to evaluate the amplitude of circles. The benefits of HTDh compared to the other variables lies in its potential validity despite its relative simplicity in assessment. Because computing HTDh requires only the positional data of the head and toes, it may have greater practical applications as an evaluative tool in gymnastics. Keywords: gymnastics, rotation, quality, evaluation, feedback, coaching, judging. INTRODUCTION "Circles" are the most basic skill on pommel horse, and they are fundamental skills due to the fact they are precursors to the development of more complex skills. Coaches focus much attention on the technical quality of the fundamental skill because of its importance in the effective development of successful performance. As George explained in his book (George, 2010), maximising the movement amplitude is one of the keys to optimising technique and successful performance in gymnastics, and this is well applied for circles. The important aspect for the quality of circles is the amplitude of horizontal rotation. Horizontal rotation refers to phase dependent rotation of: rotation of the mass centre about the Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 supporting hands and rotation of the body itself about the mass centre (Fujihara, Fuchimoto and Gervais, 2009). This partially describes the mechanics of horizontal rotation, yet more studies will be helpful to facilitate our understanding of the motion. Among practitioners, however, it would be a clearer agreement that a gymnast should keep his body as straight as possible throughout an exercise (Karacsony & Čuk, 1998; Yoshida & Shiroma, 2001). Figure 1 provides a typical example of higher-quality and lower-quality circles, which have a relatively straight body and bent body, respectively. Kaneko (1974) asserted that ideal circles should show great amplitude characterized by a straight body and maximal possible circular trajectory of his feet. This technical claim are in line with the rules, as the Code of Points (International Gymnastics Federation, 2017) states that "ideally circles and flairs must be performed with complete extension. Lack of amplitude in body position is deducted as an individual deduction on for each element. (p. 54)" It should be noted that the Code of Points (International Gymnastics Federation, 2017) also states that "circles with a slightly hollow position are permitted. (p.54)" The question here is how slight the position would be considered permissible, and this is likely to rely on judges' subjective perception at least in the current paradigm. Front support -Entry -Rear support Exit..........Front support Figure 1. Description of typical higher-quality (top) and lower-quality (bottom) circles and phase definitions. Phases are often divided into four phases based on the relative position of the legs with respect to the pommel horse and the hand release and re-grasp. Front and rear support phases are the double-hand support phases, and entry and exit phases are the singlehand support phases. In biomechanics, some studies have objectively examined the amplitude of circles. Most common mechanical variables that can discriminate the level of expertise include hip or body angle, and the horizontal diameter of the ankles' excursion (Baudry et al., 2009; Fujihara & Gervais, 2012a). When the hip or body angles were considered, some studies computed the flexion and side flexion separately (e.g. Fujihara and Gervais, 2010). According to Baudry et al. (2009), the horizontal diameter of shoulders' excursion and the shoulder extension angle in the rear support phase (See Figure 1 for the phase definitions) also show significant differences between the skilled and the unskilled gymnasts. These mechanical variables could be helpful to objectively evaluate the quality of circles, but the complexity of computation and a time-consuming data collection procedure may limit a practical application in a daily training. As a simpler variable, Fujihara and Gervais (2013) pointed out that the head Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 position particularly in the rear support phase is strongly correlated with the scores given by the official gymnastics judges. In fact, this variable showed high correlations with other amplitude variables introduced above: a body flexion angle, the horizontal diameter of shoulders' excursion and the shoulder extension angle in the rear support. Intriguingly, the horizontal diameter of ankles' excursion was the only variable that did not show significant correlation with the head position in the rear support. Because all investigated variables were more or less correlated with the scores by the judges, the motions of head and feet may contain unique information for score prediction whereas other measures such as a body or shoulder angle could be regarded as redundant with the information about the head motion. Based on these previous studies, we hypothesized that using these two body evaluate the amplitude of circles in a simpler manner. By taking into account the trajectory of the feet in addition to the position of the head segment, which has redundant information on body and shoulder angles, the distance between the head and the centre of the toes (Head-Toe Distance: HTD) can represent much about the quality of circles despite its simplicity (Figure 2). It appeared to be valid from both a practical viewpoint and a theoretical viewpoint of 'straight body' as supported by the previous research. The purpose of this study was to investigate the possible use of HTD as a simple measure for evaluating the amplitude of circles on pommel horse. Such a simple objective measure could be beneficial for evaluation of the performance, assisting coaches and athletes during training and return to best play following injury. Also, this simple measure could potentially provide parts, namely head and feet, or toes in order to differentiate whether or not toes are pointed, might be a plausible method to additional assessment information for judges in a practical setting. HTDh Head position Body flexion Shoulder extension Figure 2. Illustrations of Head-Toe Distance (HTD), its horizontal version (HTDh), and the other amplitude variables computed. Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 METHODS Data collected from successful previous studies were employed (Fujihara & Gervais, 2012b) and extended the analysis. Fourteen national-level and four international-level male gymnasts participated in the experiment. All 18 gymnasts were able to consistently perform 20 consecutive circles on two handles of a pommel horse. The mass, height, and ages of the gymnasts were 47.7 ± 10.8 kg, 1.55 ± 0.11 m, and 16.2 ± 3.6 years. They had 9.4 ± 2.9 years of experience in competitive gymnastics and trained 20.3 ± 3.5 hours per week at the time of data collection. Written informed consent was obtained of all the gymnasts and the guardians of under-18 gymnasts. Prior to the start of this project, ethical approval was obtained through the institutional research ethics review board. A no-leg pommel horse was placed in the laboratory, and the kinematic data were collected using a Qualisys motion capture system with 13 cameras placed around the pommel horse (ProReflex MCU 240, f6 lens, Qualisys AB, Sweden). According to the results of the inter-marker-distance accuracy tests, which were performed before and after each data collection, the standard deviation and the range of the 750.2 mm-wand length were at most 1.3 mm and 8.3 mm, respectively. The centre of the top surface of the pommel horse was defined as the origin of the laboratory coordinate system, and the X-, Y-, and Z-axes were the horizontal lines along the long and short axes of the horse and the vertical line through the origin. The positive directions of the X-, Y-, and Z-axes were rightward, forward, and upward with respect to the direction of the performance (Figure 2). After general and event-specific warm-ups, the participants were fitted with retro-reflective markers on the anatomical landmarks suggested by de Leva (1996) and performed three sets of 10 circles on the pommel horse. For the full description of the marker placement and the procedure of the anatomical calibrations, please refer to Fujihara and Gervais (2012b). Three-dimensional (3-D) kinematic data were recorded at 100 Hz. For each set of 10 circles, 7 circles (3rd - 9th) were used so that the individual mean data were computed from the data of 21 circles. The 3-D coordinate data were smoothed using a fourth-order Butterworth digital filter at the optimal cut-off frequencies (3.0 Hz -12.2 Hz) determined by an automatic algorithm (Yokoi & McNitt-Gray, 1990). The head segment was defined as a line from a vertex (top of head) to the centre of right and left gonions, and its mass centre was estimated as a point at 59.76% from the vertex (de Leva, 1996). Hip joint centres were estimated using Halvorsen's algorithm (2003), and all other joint centres were estimated as the centres of two markers attached on the surface of each joint. HTD was computed as the 3-D distance between the head mass centre and the centre of two markers attached on the right and left acropodions (Figure 2). The head mass centre should reflect the position of the segment better than a vertex. HTD was also computed on a horizontal plane (HTDh), and both HTD and HTDh were normalized by the body height of each gymnast (unit: %BH). As representative kinematic variables for amplitude evaluation of circles, the following variables were also considered: horizontal diameters of shoulder and ankle trajectories (shoulder diameter, ankle diameter), a body flexion angle, a shoulder extension angle in the rear support, and the head position in the rear support position (Fujihara & Gervais, 2012a; b; 2013). The diameters and angles were computed by following Grassi et al. (2005) and Fujihara and Gervais (2010), respectively. The illustrations of these variables are shown in Figure 2, but more detailed computational descriptions can be found in Fujihara and Gervais (2012b). Four internationally accredited judges Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 scored the video-recorded circles. A perfect score was set at 10.0 and deductions were applied in step of 0.1 according to technical faults or execution errors. Then, the average of four scores was determined as the final score. The intra-class correlation coefficient, computed as an estimate of the inter-judge reliability, was 0.944. Pearson's product-moment coefficient of correlation was considered among the scores and all computed variables. The normality of the data was checked using the Kolmogorov-Smirnov test. The experiment-wise error rate was set at p < 0.05, and the Holm's correction (Holm, 1979) was applied for multiple univariate statistics (Knudson, 2009). The purpose of this study was to explore the possible use of HTD as a simple measure in relation to the quality of circles, instead of to find the best way to predict a score given by the judges participated in this particular study. Therefore, stepwise multiple regression analysis was conducted by a two-block method: first block took HTD and HTDh, and then the second block took the all other variables computed as potential predictors. IBM SPSS statistics v.23 was used for the statistical computations. RESULTS Table 1 shows the scores given by the judges and the all computed variables for each individual gymnast. According to the results of the Kolmogorov-Smirnov tests, the distributions of all variables analyzed were not significantly different from a normal distribution. Also, there was no outlier that largely deviated from each variable's mean (< 3 standard deviation). The overall averages of HTD and HTDh for all gymnasts were 88 ± 3 %BH and 75 ± 2 %BH, respectively. The gymnast who had the best score out of 18 gymnasts exhibited the greatest HTD (94 %BH) and HTDh (80 %BH) as the average of 21 circles analysed (Table 1). See Figure 3 for the time-series change in HTD and HTDh during circles for each gymnast. Both HTD and HTDh tended to decrease around the rear support phase where the difference in the head position was found in the previous study (Fujihara & Gervais, 2013). However, the higher-scored gymnasts, who had scores greater than 9.0 out of 10.0, showed less decrease in both HTD and HTDh, maintaining a relatively high value throughout a circle (Figure 3). The average standard deviation of HTD and HTDh for each gymnast (21 circles) was 0.43 ± 0.16 %BH and 0.56 ± 0.18 %BH, so the intra-gymnast variability was smaller than the inter-gymnast variability. Figure 4 plots the standard deviations of HTD and HTDh for each gymnast against the scores. The higher-scored gymnasts tended to show smaller variability, implying more consistent performance. Table 2 displays the correlations among the scores and all the amplitude variables computed. The average HTD and HTDh were both significantly correlated with the scores given by the official gymnastics judges, but HTDh showed a stronger correlation than HTD (HTD: r = 0.556, p = 0.017, HTDh: r = 0.735, p = 0.001). As a result of the stepwise multiple regression analysis, two models showed statistical significance (Table 3). The first model included HTDh alone as a predictor (adjusted R2 = 0.511). The second model included HTDh and HTD, increasing the adjusted R2 to 0.662. After these two variables were entered at the first step, no other amplitude variables were entered at a statistically significant level. Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 Table 1 Scores and the performance variables for each individual gymnast. Gym -nast Score HTD (%BH) HTDh (%BH) Head position (m) Body flexion (°) Shoulder extension (°) Shoulder diameter (%BH) Ankle diameter (%BH) 1 9.675 ± 0.340 94 80 -0.15 2 34 30 101 2 9.575 ± 0.544 91 77 -0.15 -1 26 31 100 3 9.375 ± 0.556 90 77 -0.11 9 20 30 99 4 9.350 ± 0.387 89 77 -0.11 9 28 31 98 5 9.300 ± 0.594 89 76 -0.17 16 26 29 100 6 9.300 ± 0.622 90 77 -0.17 11 15 29 100 7 9.100 ± 0.898 92 78 -0.20 -2 33 32 95 8 9.100 ± 0.707 88 75 -0.09 13 22 29 100 9 8.975 ± 0.822 87 73 -0.05 24 16 27 98 10 8.975 ± 0.608 88 76 -0.07 16 17 28 103 11 8.750 ± 0.686 91 77 -0.18 15 27 31 96 12 8.575 ± 0.939 89 76 -0.15 3 24 29 98 13 8.425 ± 0.911 86 75 -0.06 15 18 28 98 14 8.425 ± 0.709 84 71 -0.05 22 7 26 95 15 8.300 ± 1.030 84 71 -0.01 35 10 26 96 16 7.950 ± 0.526 88 74 -0.05 25 9 28 94 17 7.300 ± 0.963 87 73 -0.08 24 21 29 95 18 6.650 ± 1.109 87 72 -0.04 35 14 27 94 Aver -age 8.728 88 75 -0.10 15 20 29 98 SD 0.798 3 2 0.06 11 8 2 3 Figure 3. Change in HTD and HTDh during a single circle. Each line represents the average of 21 circles by each gymnast. Higher-scored (>9.0) - and lower-scored (< 8.5) gymnasts are shown in black and grey lines, respectively. The grey broken lines are the gymnasts whose scores were between 9.0 and 8.5. The thickest black line shows the highest-scored gymnast (9.675). The gymnasts with higher scores demonstrate relatively higher HTD and HTDh throughout a whole circle, particularly from the rear support. Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 M -T3 3 c > 'S 10.0 9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 r = 0.55 (p < 0.05) 0.0 0.2 0.4 0.6 SD of HTD (%BH) 0.8 1.0 10.0 9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 0.0 r = 0.64 (p < 0.01) 0.2 0.4 0.6 0.8 SD of HTDh (%BH) 1.0 Figure 4. The relationships between the scores given by human judges and the standard deviation of HTD (left) and HTDh (right) among 21 circles (7 circles x 3 sets). Table 2 Correlations among the scores and the amplitude variables. Variables Critical p Sig. Body flexion 0.737 0.000 0.007 HTDh 0.735 0.001 0.008 Ankle diameter 0.725 0.001 0.010 Head position -0.576 0.012 0.013 HTD 0.556 0.017 0.017 Shoulder extension 0.517 0.028 0.025 Shoulder diameter 0.498 0.035 0.050 Table 3 The results of stepwise multiple regression analysis. Partial Correlation (Excluded variables) Model Predictors Adjusted R2 F Sig. Head Body Shoulder Shoulder HTD Position flexion extension diameter Ankle diameter 1 HTDh 0.511 18.78 0.001 -0.593 0.059 0.295 -0.172 -0.297 0.550 2 HTDh, HTD 0.662 17.65 0.000 - -0.24 0.329 0.036 -0.006 0.304 DISCUSION Providing simple, scientifically rigorous and ecologically valid methods to assess the quality of skill is a desirable tool for coaches, sports judges, and scientists. This study investigated the use of HTD, a simple measure using head and toes to evaluate the quality of circles, finding that its horizontal component, HTDh, is a potential variable. The results shown in Figure 3 supported the possible use of HTDh both on individual and group levels. The higher-scored gymnasts showed less decrease in HTDh during and after the rear support phase, resulting in the greater Science of Gymnastics Journal 189 Science of Gymnastics Journal r Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 HTDh on average (Table 1). HTD also showed a strong relationship with the scores and the other variables but not as strong as that of the HTDh. When both HTDh and HTD were used to predict the scores, more variance was significantly explained (Table 3). Whilst there are more thorough ways to evaluate the quality of circles e.g. using multiple biophysical variables to reflect different aspects, but HTDh seems to be a good option as a single variable to evaluate the amplitude of horizontal rotation. Recall that in this study the judges followed two rules: that a perfect score was 10.0, and that a deduction was applied in step of 0.1 according to technical faults or execution errors. The highest score seen among the participants was 9.675 (Table 1), so all of them seemed to have room to improve their performance to some extent. In fact, some judges provided their comments on the performances in addition to the score. Based on their comments, a judge appears to look at body extension, amplitude, rhythm, leg form errors, consistency, and the timing of hip turns. The scores given by the judges in this study should be considered to be more comprehensive and therefore more complicated than just focusing on the amplitude. In Figure 4, a few of the lowest-scored gymnasts showed relatively greater deviation from the regression line, implying that the judges more than likely found multiple technical errors, including the amplitude, in those performances. As shown in Figure 4, the lower-scored gymnasts demonstrated less consistent performances. Looking to future studies, an examination of the relationship between the amplitude variables with respect to human-judged evaluation that is concentrated only on the amplitude would be interesting. Using HTDh as an objective measure for the amplitude may become more attractive when its computational simplicity is taken into account. For instance, the body flexion angle was slightly better than HTDh for predicting a score (Table 2), but this variable used in this study as well as the previous studies (Fujihara & Gervais, 2010; 2012) required the 3-D angular computations in a couple of moving coordinate systems. The results of 3-D angular computations were often influenced by the definition of angles as well as the way to determine the joint centre. HTDh, on the other hand, required only the 3-D positional data of the two points and a simple computation of the distance between them. It is much simpler and easier to quantify the amplitude as a measure of the performance quality. Computing the head mass centre using Grassi et al's method, which computes the horizontal diameter of ankle or shoulders trajectory, also needs only positional data to quantify performance quality (Grassi et al., 2005). In their method, the average position throughout a whole circle is regarded as the centre of rotation, and the radius or diameter is calculated based on this rotational centre at each time point. It is, therefore, impossible to evaluate the amplitude at any single time frame independently. In contrast, HTDh does not need any reference point for its computation. It is only the distance between the two points at any time. Furthermore, there is no need for the identification of the phases. Because the shoulder extension angle and the head position as the amplitude variables were expected to show significant differences during the rear support phase, identifying a phase is a necessary requirement to compute these variables. While some may argue that sports like gymnastics need objective scores to prevent issues associated with human qualitative assessment, others may argue that the quality of a performance in such sports should not or cannot be evaluated in any quantitative manner. This is mainly a matter of validity. The question is not whether it is qualitative or quantitative but whether it is valid or invalid. Human perception tends to be influenced by Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 various factors. For example, Plessner and Schallies (2005) reported the significant influence of viewpoint when the gymnastics judges were asked to evaluate the Cross on rings. Judges in artistic gymnastics are seated in assigned locations around the apparatus when judging performances at competition (Article 5.6, p18, International Gymnastics Federation, 2017), implying people tacitly accept a possible influence of a view angle. Considering the plane of motion during circles on pommel horse, the amplitude of horizontal rotation would be best evaluated from above. Figure 5 may help us to accept the possible influence of the view angle on Front view Gymnast A our impression of the performance. Although recent TV broadcasts often include video from above, in the current rules, no judge on pommel horse may evaluate from above. An aerial view is also unlikely to be available in regular training facilities. As discussed in Fujihara, Yamamoto, and Fuchimoto (2017), objective information could become more useful to supplement subjective feedback or evaluation if such objective information with good validity and reliability is provided with an easy and immediate manner at a low cost. The immediate feedback of the HTDh could be of very practical benefit to gymnasts and coaches. Diagonal view Gymnast B Figure 5. Descriptions of the possible different impression caused by a different view position. For each gymnast A or B, two figures shows the rear support position at the same time from a different viewpoint. The difference in body alignment, especially a piked body in gymnast B, is less clear in the front view. It should be noted that the original videos for tracing these figures were not those that were recorded for the current study. Therefore the models for these figures are do not match with any individual shown in Table 1. This simplicity may contribute to a more practical application of biomechanical knowledge in a regular training or possibly in a competition. To date in the gymnastics community, the complexity of typical biomechanical procedures have limited the practical use of such an objective information for coaching in a regular training session or judging in a competition. Nevertheless, more recent technology may change the situation in the near future. Mr. Morinari Watanabe, the current president of the International Gymnastics Federation, Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 presented his plan to develop a judging support system in collaboration with a Japanese company, Fujitsu (Liubov, 2017). Fujihara (2017) introduced the potential usage of a Kinect device to identify the top of the head and the tip of toes during circles on pommel horse, as a non-invasive approach to provide immediate feedback in a regular training session. Although the Kinect system captured the positions of the top of the head, not the head mass centre, to compute HTDh, the possible difference should be within an acceptable range in a practical setting. When we computed HTDh with a vertex instead of the head mass centre, the Pearson's correlation coefficient was still high (r = 0.69). On a horizontal plane, the top of the head would be closer to the head mass centre than a vertex. To use HTDh as an amplitude measurement, there are several things to be considered. First, what would be the best use of HTDh was not considered in the current study. Only the average of HTDh during a whole circle was computed, but there might be a more valid way to evaluate a performance. Because especially in the front support position, it is less persuasive to argue that the greater HTDh always means the better performance. Therefore, setting a certain benchmark value for amplitude could be a possible option. Let us say that 80% of body height is the standard, and then a point is deducted if HTDh is lower than this value thus indicating lack of amplitude. Such an evaluation method was tested, but the correlation with the scores was very similar to the case with mean HTDh (r = 0.73) with the current data set. Second, using multiple variables in addition to HTDh could improve the score prediction because the ankle diameter still seemed to contain unique information, which was not reflected in HTDh. One possible explanation for this is the presence of centrifugal force. Fujihara (2016) presented that skilled gymnasts are better at maintaining a higher centrifugal force during a transitional phase than their unskilled counterparts. A greater centrifugal force contributes to a horizontal rotation of the mass centre that is dynamically balanced on a higher plane, resulting in a greater diameter of the distal points (feet and toes). This detail may not be fully reflected in HTDh, therefore accompanying measures could improve the validity of the evaluation. We have to acknowledge that the results of this study were based on relatively homogenous samples of the gymnasts and the scores of only four judges. Finding the best way of score prediction was beyond the purpose of this research. Additionally, a different aspect of performance quality, such as consistency, will be a task for future study. CONCLUSIONS Previous findings led us to hypothesize that the distance between the head and toes, HTD and HTDh for its horizontal version, could be a more practical and simpler tool to evaluate the amplitude of circles. In this study, we investigated HTD and HTDh by examining the relationships between these novel variables and the human-judged score as well as the other amplitude variables used in the literature. The results supported the use of HTDh rather than HTD or any other amplitude variable as a single variable to objectively evaluate the amplitude of circles on pommel horse. In addition to the potential validity as an objective measure, its simplicity in concept and an actual process to obtain the data may provide us more opportunities to apply it in a practical setting. As such this grounded scientific approach with high ecological validity presents practical applications and warrants the further exploration of the use of HTDh. Integrating an advanced technology and a novel idea may enhance the quality of information to the sport. Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 REFERENCES Baudry, L., Sforza, C., Leroy, D., Lovecchio, N., Gautier, G., & Thouvarecq, R. (2009). Amplitude variables of circle on the pedagogic pommel horse in gymnastics. Journal of Strength and Conditioning Research / National Strength and Conditioning Association, 23(3), 705711. de Leva, P. (1996). Adjustments to Zatsiorsky-Seluyanov's segment inertia parameters. Journal of Biomechanics, 29(9), 1223-1230. Fujihara, T. (2017) Using a kinect device to evaluate amplitude of horizontal rotation on the pommel horse. Proceedings of the 35th Conference of the International Society of Biomechanics in Sports, Cologne. Germany. Fujihara, T. , Fuchimoto, T., & Gervais, P (2009) Biomechanical analysis of circles on pommel horse. Sports Biomechanics, 5(1), 22-38. Fujihara, T., Yamamoto, E., & Fuchimoto, T. (2017). Run-up velocity in the gymnastics vault and its measurement. Japan Journal of Physical Education, Health, and Sport Sciences, 62, 435-453. Fujihara, T., & Gervais, P. (2010). Kinematics of side and cross circles on pommel horse. European Journal of Sports Sciences, 10(1), 21-30. Fujihara, T., & Gervais, P. (2012a). Circles on pommel horse with a suspended aid: Influence of expertise. Journal of Sports Sciences, (30)6, 583-589. Fujihara, T., & Gervais, P. (2012b). Circles on pommel horse with a suspended aid: Spatio-temporal characteristics. Journal of Sports Sciences, (30)6, 571-581. Fujihara, T., & Gervais, P. (2013) Head positions are related to the performance quality of circles on pommel horse. Proceedings of the 31st Conference of the International Society of Biomechanics in Sports, Taipei. Taiwan. Fujihara, T., Yamamoto, E., & Gervais, P. (2015). Toward an ideal performance of circles on pommel horse— centrifugal force and mass-centre velocity—. Proceedings of the 33rd International Conference on Biomechanics in Sports, Poitiers, France. George, G.S. (2010). Championship gymnastics: Biomechanical techniques for shaping winners. Carlsbad, CA: Designs for Wellness Press. Grassi, G., Turci, M., Shirai, Y. F., Lovecchio, N., Sforza, C., & Ferrario, V. F. (2005). Body movements on the men's competition mushroom: a three dimensional analysis of circular swings. British Journal of Sports Medicine, 39(8), 489-492. Halvorsen, K. (2003). Bias compensated least squares estimate of the center of rotation. Journal of Biomechanics, 36(7), 999-1008. Holm, S. (1979). A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics, 6, 65-70. International Gymnastics Federation. FIG Men's Technical Committee (Ed.) (2013). Code of Points. Retrieved from http://www.fedintgym.com/rules/docs/06-code/01-mag/codemag0701-efs.zip International Gymnastics Federation. FIG Men's Technical Committee (Ed.) (2017). Code of Points. Retrieved from http://www.fig- gymnastics.com/publicdir/rules/files/en_M AG CoP 2017-2020.pdf Kaneko, A. (1974). Coaching of artistic gymnastics Tokyo: Taishukan. Karacsony, I., & Čuk, I. (1998). Pommel horse exercises: methods, ideas, curiosities, history Ljubljana: University of Ljubljana and Hungarian Gymnastics Federation. Knudson, D. (2009). Significant and meaningful effects in sports biomechanics research. Sports Biomechanics, 8(1), 96104. Liubov, B. (2017). Robot judging is coming to gymnastics. Retrieved Date Accessed, 18th September, 2017 from Science of Gymnastics Journal 189 Science of Gymnastics Journal Fujihara T., Gervais P., Irwin G.: HEAD-TOE DISTANCE AS SIMPLE MEASURE TO... Vol. 11 Issue 2: 189 - 200 http://eng.gymnovosti. com/robot-judging-is-coming-to-gymnastics/. Plessner, H., & Schallies, E. (2005). Judging the cross on rings: a matter of achieving shape constancy. Applied Cognitive Psychology, 19(9), 1145-1156. Yokoi, T., & McNitt-Gray, J. L. (1990). A threshold to determine optimum cutoff frequency in automatic data smoothing using digital filter. In American Society of Biomechanics (Ed.), Proceedings for the 14th annual meeting of the American Society of Biomechanics (pp. 209-210). FL: University of Miami. Yoshida, K., & Shiroma, A. (2001). Pommel horse. In Men's Artistic Gymnastics Department of Japan Gymnastics Association (Ed.), Training manual for male junior gymnasts (pp. 50126). Wakayama: Minakuchi Corp. Corresponding author: Osaka University of Health and Sport Sciences - School of Health andSport Sciences 1-1, Asashirodai Kumatori-cho Sennan-gun Osaka 590-0496 Japan T: +81 72 479 3115 E-mail: fujihara@ouhs.ac.jp Science of Gymnastics Journal 189 Science of Gymnastics Journal Milcic L., Zivcic K., Krsticevic T.: DIFFERENCES IN VAULT RUN-UP VELOCITY... Vol. 11 Issue 2: 201 - 207 DIFFERENCES IN VAULT RUN-UP VELOCITY IN ELITE GYMNASTS Lucija Milcic, Kamenka Zivcic, Tomislav Kristicevic University of Zagreb, Faculty of Kinesiology, Croatia _Original article Abstract The aim of this study was to compare differences in run-up velocity between Handsprings, Tsukahara and Yurchenko entry on vault. A sample consisted of 48 jumps performed on vault, 19 Handsprings, 17 Tsukahara and 12 Yurchenko entry on vault. Data were collected on a World Cup competition held in Osijek, 2017. Run-up velocity was measured by speed radar gun (Stalker ATS, S PRO II). Descriptive statistic was calculated for all variables and differences in run-up velocity were determined by one-way ANOVA and Bonferroni post-hoc test at the level of statistical significance at p<.05. Average run-up velocity at Handspring entry was 8,06 m/s, Tsukahara, 8,06 m/s and Yurchenko entry on vault table was 7,66 m/s. ANOVA showed that exist statistical significant differences in run-up velocity between handspring and Yurchenko and between Tsukahara and Yurchenko entry. The results of this study indicated that different entry on vault table has different run-up velocity, which will help coaches and scientists to improve the vault technique. Keywords: artistic gymnastics, vault, velocity, run-up. INTRODUCTION The vault is an apparatus in artistic gymnastics characterized by complex and very short (not much more than 7 seconds in average) dynamic movements (Čuk & Karacsony, 2004). Different from other apparatus on which gymnasts performing exercise lasting for ninety seconds, jumps on vault lasting up to five to six seconds. Gymnastic elements at the vault are classified on the structural characteristics of the jumps, which, together with the development of artistic gymnastics, have been further refined and supplemented (Živčic Markovic & Krističevic, 2016). Velocity is defined as a function of time where measures the distance of the object travelled through time, and measurement unit is in meters per second (m/s). As the run-up velocity did not investigate on competitions since 2012., we wanted to see if there are some changes considering that the elements on all apparatus are more difficult to perform. Entry on the apparatus can be in the forward, backward and sideward direction to the vault. The most common entry on vault are Handsprings, Tsukahara/Kasamatsu and Yurchenko. Handsprings jumps are executed after runup and take-off from the springboard in the forward direction with face to the vault, after that hands are placed in shoulder-width on the vault table. From that position hands do take-off from the vault table and performing the salto with different rotations. Yurchenko vaults are consisted Science of Gymnastics Journal 201 Science of Gymnastics Journal Milcic L., Zivcic K., Krsticevic T.: DIFFERENCES IN VAULT RUN-UP VELOCITY... Vol. 11 Issue 2: 201 - 207 of run-up and performing round-off which ends with a take-off on the springboard. After that gymnasts performing backward handspring with different salto from tucked to stretched salto with turns. Tsukahara jumps are consisted of run-up and take off from the springboard and performing handsprings with turn for 90° or 180° and take off from the apparatus in backward salto. Vault consist of seven phases: 1. run-up; 2. jumping on the springboard; 3. springboard support; 4. first flight phase; 5. table support; 6. second flight phase and 7. landing (Prassas & Gianikellis, 2002; Prassas, Kwon, & Sands, 2006; Čuk & Karacsony, 2004; Takei, Dunn, & Blucker, 2007; Ferkolj, 2010; Atikovic & Smajlovic, 2011). Each phase depends on previous, so it is important that execution of all seven phases is without mistakes. The most investigated group of jumps on the vault are Handsprings (45%), Yurchenko (10%) and Tsukahara (5%) (Fernandes, Carrara, Serrao, Amadio & Mochizuki, 2016). There is a little investigation that is conducted on the new vault table on the competition considering that the technique of the elements constantly progresses and today's gymnasts performing much more elements with high values and complexity. In artistic gymnastic tendency of gymnasts and coaches are to perform the hardest elements which means that those elements have highest difficulty value, and also, they are much more complex for execution than elements with lower difficulty value. Run-up velocity have been studied on competition (Van der Eb, et al., 2012; Naundorf, Brehmer, Knoll, Bronst, & Wagner, 2008; Takei, et al., 2000; Sands, 2000; Krug, Knoll, Kothe, & Zocher, 1998) and on the training (Bradshaw, Hume, Carlton, & Aisbett, 2010; Brehmer, 2011). Generally, is accepted that the speed of the run-up and take off from the springboard, linear and angular momentum are more important than contact with the vault table (Prassas, et al., 2006). After the introduction of the new vault table, the run-up velocity has significantly increased since 1997. year in both genders, excepting Yurchenko jumps where men were faster than women, so new vault table does not have the influence of the run-up velocity (Naundorf, et al., 2008). Velocity of run-up has been stabilized between 1997 and 2010, because of physical limitation or it is optimally run-up velocity for the type of the vault (Van der Eb, et al., 2012). Positive correlations between take-off velocity and final score indicate that increase in run-up velocity will influence on take-off velocity (Van der Eb, et al., 2012). This is the explanation of all phases at the vault, which have to be connected. If run-up is slow, the next phase of the vault or difficulty value of the jump will be lower. Each gymnast will develop optimal velocity by running which depends on the type and complexity of jump. The gymnast should try to attain the highest velocity that can control (Sands, 2000). The run-up velocity is related to the complexity of the jump, for example, run-up velocity is lower at Yurchenko jumps than at Handsprings jumps, but the velocity decreases in the jumps where the rotation is reversed in the second stage of the flight (Prassas & Gianikellis, 2002). Veličkovic, Petkovic, & Petkovic, (2011) investigated run-up velocity with 3D kinematics analyses and Optojump system, and results have shown that there are differences in run-up velocity between elite gymnasts (in the middle of the run-up was recorded a decrease of velocity), but in the average gymnast run-up velocity constantly grows. Čuk, Bricelj, Bučar, Turšič, & Atikovic (2007) investigated the connection between the start value of the vault and run-up velocity in top level male artistic gymnasts, and results have shown that there is a correlation between run-up velocity and the start value of that jump. In sports biomehanics various devices have been used for velocity measurement. Laser and radars mostly are used for measuring running velocity. Harrison, Science of Gymnastics Journal 201 Science of Gymnastics Journal Milcic L., Zivcic K., Krsticevic T.: DIFFERENCES IN VAULT RUN-UP VELOCITY... Vol. 11 Issue 2: 201 - 207 Jensen, & Donoghue, 2005 investigated validity and reliability of laser measuring system compared with the reliability of video based system during running, and have concluded that laser system are validated and reliable instrument for velocity measurements. In artistic gymnastics to determine run-up velocity are used lasers (Krug, et al., 1998; Naundorf, et al., 2008). The radar system is also used because of very simple usage (Sands, 2000). Recently are used OPTO-TRACK system which has optical sensors placed along the whole track (Velickovic, et al., 2011). Motor skills that are important for performing vault jumps are strength, especially explosive power for take-off from legs and hands, and velocity. Jumps on vault require an explosive power of upper and lower extremities. The run-up velocity is transferred to a springboard, allowing the gymnast to perform a successful jump. The aim of this paper was to determine the differences in the run-up velocity during the performance of the different groups of vaults jumps on the competition. METHODS The sample consisted of 48 jumps performed on the vault of which 19 were Handsprings, 17 Tsukahara and 12 Yurchenko entry to the vault. Data were collected in competition, at the World Cup in Osijek, 2017 (MAG and WAG Artistic Gymnastics). Run-up velocity was measured by Stalker ATS, S PRO II (Applied Concepts, Inc., Texas, USA), from the beginning of the run-up to the moment of take-off, i.e. the last step before the springboard. The device was placed behind the mat for landing. Stalker radar has a speed range of 0.2 m/s to 18.0 m/s -from below 1 mph (miles per hour) to over 40 mph - with an accuracy of ±0.1 m/s. High speed is 150 mph (241 km/h, 130 knots, 67 m/s). When it was compared with photocells which are the gold standard for speed measurement, validity was r2 = 0.99, p\0.01 (Chelly & Denis, 2001; di Prampero, Fusi, Sepulcri, Morin, Belli & Antonutto, 2005; Haugen & Buchheit, 2016; Morin, Jeannin, Chevallier, & Belli, 2006). Run-up velocity was measured in km/h, and converted into m/s, with the mathematic formula: km/h x 1000/3600. Variable Hnd was used for a Handspring, Tsuk for Tsukahara and Yurc for Yurchenko vault. This study was approved by Ethical Committee of Faculty of Kinesiology. Statistica 12 was used for data analysis. Normality of distribution was determined by Kolmogorov-Smirnov test. Variables were normally distributed. Basic descriptive parameters were calculated for all variables, and the differences of run-up velocity between groups of vault jumps were determined by the one-way ANOVA and the Bonferroni post-hoc test. Level of statistical significance was set at p<0.05. RESULTS Basic descriptive parameters of measured variables are shown in Table 1. Results of One-way ANOVA for Velocity of measured variables are shown in Table 2. Results indicated that there were statistical significant differences between run-up velocities of vault entry p<.01. In Table 3 are shown results of Bonferroni Post Hoc test for variables group of vaults. There is a statistical significant difference p<0.01 difference in run-up velocity between the Handsprings and Yurchenko entry, and between Tsukahara and Yurchenko entry on vault. Science of Gymnastics Journal 201 Science of Gymnastics Journal Milcic L., Zivcic K., Krsticevic T.: DIFFERENCES IN VAULT RUN-UP VELOCITY... Vol. 11 Issue 2: 201 - 207 Table 1 Descriptive parameters of run-up velocity on different vault elements. Variables Valid N Mean Min Max Std.Dev. Hnd 19 8.06 7.40 8.90 0.36 Tsuk 17 8.06 7.08 8.68 0.41 Yurc 12 7.66 7.25 7.93 0.26 Note. Hnd: run-up velocity of Handspring entry on vault; Tsuk: run-up velocity of Tsukahara entry on vault; Yurc: run-up velocity of Yurchenko entry on vault; Valid N: number of samples; Mean: average values; Min: minimum values; Max: maximum values; Std.Dev.: Standard Deviation. Table 2 One -way ANOVA analysis of velocity. Dependent Multiple Multiple Adjusted SS df MS SS df MS F p Variable R R2 R2 Model Model Model Residual Residual Residual Velocity 0.45 0.20 0.17 1.47 2 0.74 5.83 45 0.13 5.68 0.01* m/s Note. * level of significance p<0.05. Table 3 Bonferroni Post Hoc test. Group of vaults {1} {2} {3} 8,06 8,06 7,66 1 1.00 0.01* 2 1.00 3 0.01* 0.01* 0.01* Note. {1}: Handsprings entry on vault; {2}: Tsukahara entry on vault; {3}: Yurchenko entry on vault; * level of significance p<0.05. DISCUSSION The results showed that the run-up velocity for Handsprings entry to the vault was 8.06 m/s, Tsukahara 8.06 seconds and Yurchenko entry to the vault was 7.66 m/s. Yurchenko entry had a slightly lower runup velocity than the Tsukahara entry. Investigations of run-up velocity have shown differences between vault type or entry on vault table, depending on weather are performing Yurchenko, handspring or Tsukahara vault (Sands, 2000; Veličkovic, et al., 2011; Naundorf, et al., 2008; Dolenec, Čuk, Karacsony, Bricelj, & Čoh, 2006). In the last decade, vault run velocity has changed. Those changes are influenced by new vault table and each Olympic cycle new Code of Points. Naundorf, et al (2008) have investigated run-up velocity and concluded that the run-up velocity has been significantly increased since 1997. The reason for that is in the precision of placing the hands in the first flight phase and elements which will be performed in the second flight phase. Investigation by Sands (2000; 2002), have shown that Yurchenko vaults have slower run-up velocity than Handsprings and Tsukahara vaults, because of coming with back to the apparatus and the precision of the roundoff performance before the vaulting table. These results are similar to our results. ANOVA has shown that there are significant differences in the run-up velocity between the Handsprings and Yurchenko entry, and between Tsukahara and Yurchenko entry on vault. Science of Gymnastics Journal 201 Science of Gymnastics Journal Milcic L., Zivcic K., Krsticevic T.: DIFFERENCES IN VAULT RUN-UP VELOCITY... Vol. 11 Issue 2: 201 - 207 Handsprings and Tsukahara entry have a faster run-up velocity than Yurchenko entry to the vault for better visual control of devices and springboard which allows the gymnast greater precision of entrance at the springboard and vault table and therefore a faster run-up. The Handsprings group requires a larger amplitude in the second flight time and is considered more difficult to perform (Farana, Uchytil, Zahradnik, Jandacka, & Vaverka, 2014). The reason is that from front handsprings they are usually performed rotated around the transverse axis of the body and this is the same with the Tsukahara vault, where the rotations are performed around the transverse axis, but the rotation is performed in backward rotation. Also, an important role in achieving the optimal run-up velocity is the number of steps for each jump. In one study (Čuk & Karacsony, 2004), investigated the speed of the jump according to the difficulty of execution of the jumps and concluded that the run-up velocity should be from 7.5 to 8.5 m/s, and for the difficult jumps of 8.5 to 9.5 m/s for jumps with double salto, run-up velocity should be over 10m/s. Some gymnasts need more steps to achieve the optimal run-up velocity. In each group of vault jumps, higher run-up velocity requires jumps of higher difficulty value and the number of rotations (Bradshaw, 2004). It is important where gymnast will place the springboard for different jumps and also where will start with run-up. For some vaults gymnasts use shorter or larger runup, it depends on own style of performance. Previous research has shown that errors occurring during the run-up and take-off are difficult to compensate in later stages of the jump (Prassas, et al., 2006). For example, if the gymnasts have a low run-up velocity and weak take-off, then they will reduce height of the jump and the inability to perform the predicted number of rotations during the second flight phase. Since we only collected this information for the optimal run-up velocity, the run-up velocity has its acceleration. The speed of the last ten steps is progressive, and the highest is in the last step at the top (elite) (9.95 m/s) and at High level gymnast is 8.57 m/s. (Velickovic, et al., 2011). In the last step, gymnasts decrease run-up velocity because of some preparatory motions for take-off (Fujihara, 2016). CONCLUSIONS In the development of the optimal runup velocity, it is necessary to pay attention in the training process that the maximum speed is reached before the contact with the springboard, i.e. that it has not decreased in the last steps. Today's popularity of performing double or triple rotations in the second flight phase is huge. Gymnasts trying to perform as high as possible difficulty values of jumps, so they can get better grades of referees and make competition attractively for watching. Considering the importance of investigations of run-up velocity and giving the feedback about key performance parts to coaches, further research should be more detailed to clarify the run-up velocity and put them in relations with other kinematical parameters especially on competitions. REFERENCES Atikovic, A., & Smajlovic, N. (2011). Relation between vault difficulty values and biomechanical parameters in men's artistic gymnastics. Science of Gymnastics Journal, 3(3), 91-105. Bradshaw, E. (2004). Target-directed running in gymnastics: a preliminary exploration of vaulting. Sports Biomechanics, 3(1), 125-144. Bradshaw, E., Hume, P.A., Carlton, M. & Aisbett, B. (2010). Reliability and variability of day-to-day vault training measures in artistic gymnastics. Sports Biomechanics, 9, 18. Science of Gymnastics Journal 201 Science of Gymnastics Journal Milcic L., Zivcic K., Krsticevic T.: DIFFERENCES IN VAULT RUN-UP VELOCITY... Vol. 11 Issue 2: 201 - 207 Brehmer, S.N. (2011). Age-related development of run-up velocity on vault. Science of Gymnastics Journal, 3. Chelly, S.M., & Denis, C. (2001). Leg power and hopping stiffness: relationship with sprint running performance. Medicine and Science in Sports and Exercise, 33(2), 326-333. Čuk, I., Bricelj, A., Bučar, M., Turšič, B., Atikovic, A. (2007). Relations between start value of vault and runway velocity in top level male artistic gymnastics. In N.Smajlovic (Ed.), Proceedings Book of 2rd International Scientific Symposium, Sarajevo, 2007, "NTS New Technolgies in Sport", (p. 6467). Sarajevo: Faculty of Sport and PE, University of Sarajevo. Čuk, I., & Karacsony, I. (2004). Vault: methods, ideas, curiosities, history: ŠTD Sangvinčki. di Prampero, P.E., Fusi, S., Sepulcri, L., Morin, J.B., Belli, A., & Antonutto, G. (2005). Sprint running: a new energetic approach. Journal of experimental Biology, 205(14), 2809-2816. Dolenec, A., Čuk, I., Karacsony, I., Bricelj, A., Čoh, M. (2006). Runway characteristics of vault in women gymnastics. Kalokagathia, 3(4), 127-136. Farana, R., Uchytil, J., Zahradník, D., Jandacka, D., & Vaverka, F. (2014). Differences in the key kinematic parameters of difficult Handpsring and Tsukahara vaults performed by elite male gymnasts. Science of Gymnastics Journal, 6(2), 53-61. Ferkolj, M. (2010). A kinematic analysis of the handspring double salto forward tucked on a new style of vaulting table. Science of Gymnastics Journal, 2(1), 35-48. Fernandes, S.M.B., Carrara, P., Serrao, J.C., Amadio, A.C., & Mochizuki, L. (2016). Kinematic variables of table vault on artistic gymnastics. Revista Brasileira de Educagao Física e Esporte, 30(1), 97-107. Fujihara, T. (2016). Revisiting runup velocity in gymnastics vaulting. In M. Ae, Y. Enomoto & N. Fujii, H. Takagi (Eds.), Proceedings of 34 International Conference on Biomechanics in Sports, Tsukuba, Japan (p. 593-596). Michigan: Northern Michigan University. Krug, J., K. Knoll, Kothe, T. & Zocher, H.D. (1998). Running approach velocity and energy transformation in difficult vaults in gymnastics. In: H.J. Riehle, M M. Vieten, (Ed.), ISBS '98 XVI International symposium on biomechanics in sports, (p.160-163). Germany: UVK -Universitatsverlag. Harrison, A.J., Jensen, R.L., & Donoghue, O. (2005). A comparison of laser and video techniques for determining displacement and velocity during running. Measurement in physical education and exercise science, 9(4), 219231. Haugen, T., & Buchheit, M. (2016). Sprint running performance monitoring: methodological and practical considerations. Sports Medicine, 46(5), 641-656. Morin, J.B., Jeannin, T., Chevallier, B., & Belli, A. (2006). Spring-mass model characteristics during sprint running: correlation with performance and fatigue-induced changes. International journal of sports medicine, 27(02), 158-165. Naundorf, F., Brehmer, S., Knoll, K., Bronst, A., & Wagner, R. (2008). Development of the velocity for vault runs in artistic gymnastics for the last decade. In Y.H. Kwon, J. Shim, J.K. Shim & I.S. Shin (Eds.), Proceedings of 26 International Conference on Biomechanics in Sports, Seoul, Korea (p.481-484). Germany: University of Konstanz. Prassas, S., & Gianikellis, K. (2002). Vaulting mechanics. In K. E. Gianikellis (Ed.), Proceedings of 20 International Symposium of Biomechanics in Sports, Caceres, Spain (p. 1-4). Germany: University of Konstanz. Prassas, S., Kwon, Y.H., & Sands, W.A. (2006). Biomechanical research in artistic gymnastics: a review. Sports Biomechanics, 5(2), 261-291. Science of Gymnastics Journal 201 Science of Gymnastics Journal Milcic L., Zivcic K., Krsticevic T.: DIFFERENCES IN VAULT RUN-UP VELOCITY... Vol. 11 Issue 2: 201 - 207 Sands, W.A. (2000). Vault run speeds. Technique, 20(4), 1-5. Sands, W.A., & McNeal, J.R. (2002). Some guidelines on the transition from the old horse to the new table. Technique, 22, 22-23. Takei, Y., Dunn, J.H., Blucker, E.P., Nohara, H. & Yamashita, N. (2000). Techniques used in high- and lowscoring Hecht vaults performed at the 1995 World Gymnastics Championships. Journal of Applied Biomechanics, 16, 180-195. Takei, Y., Dunn, J.H., & Blucker, E.P. (2007). Somersaulting techniques used in high-scoring and low-scoring Roche vaults performed by male Olympic gymnasts. Journal of sports sciences, 25(6), 673-685. Van der Eb, J.V.D.E., Filius, M., Rougoor, G., Van Niel, C., de Water, J., Coolen, B., & de Koning, H. (2012). Optimal velocity profiles for vault. In ISBS-Conference Proceedings Archive (Vol. 1, No. 1). Velickovic, S., Petkovic, D., & Petkovic, E. (2011). A case study about differences in characteristics of the run-up approach on the vault between top-class and middle-class gymnasts. Science of Gymnastics Journal, 3(1), 25-34. Zivcic Markovic, K., & Kristicevic T., (2016). Osnove sportske gimnastike. [Basics of gymnastics]. Zagreb: Faculty of Kinesiology, University of Zagreb. Corresponding author: Lucija Milcic Faculty of Kinesiology Horvacanski zavoj 15 10 000 Zagreb Croatia Tel and Fax num: +385 981679752 E-mail: lucija.milcic@kif.hr Science of Gymnastics Journal 201 Science of Gymnastics Journal Milcic L., Zivcic K., Krsticevic T.: DIFFERENCES IN VAULT RUN-UP VELOCITY... Vol. 11 Issue 2: 201 - 207 Science of Gymnastics Journal 201 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND PERFORMANCES IN NOVICES: INVERTING EXPLICIT AND IMPLICIT LEARNING OF SKILL-RELATED MOTOR TASKS Jonas Rohleder, Tobias Vogt German Sport University Cologne, Institute for Professional Sport Education and Sport Qualifications, Cologne, Germany _Original article Abstract In gymnastics, mainstream handstand coaching emphasizes developing an aligned rigid body configuration, frequently leaving wrist-controlled balance work to implicit learning. However, skill-related motor behavioral research suggests the wrists to primarily contribute postural control in handstands. Considering recent research on handstands revealing experience-dependent motor behavior, the present study aimed to examine motor learning effects of explicit wrist usage coaching on handstand performances in skilled and less skilled novices. Therefore, twenty-five volunteering sport students served as participants completing a three-week training intervention which solely and explicitly addressed successful wrist usage during handstand. A video-tutorial introducing participants to the wrist strategy of hand balance preceded five practical training sessions that all neglected providing explicit postural advice. Participants performed three handstands on a plane gymnastics mat prior to (pre-test) and after (post-test) completing the training intervention. Standardized video recordings of each trial allowed retrospective group assignment (skilled and less skilled novices) based on pre-test mean balance times. With this, balance times, expert assessments (postural execution and balance control strategies) and goniometric analyses of shoulder and hip joint angles served to detect practical changes in handstand performances. Enhanced balance times as well as increased scores for postural execution and balance control strategies were revealed for less skilled novices (p < .05), but not for skilled novices (p > .05). Furthermore, in both groups changes in shoulder and hip joint angles failed significance. In conclusion, present findings suggest practitioners to make entirely unexperienced handstand learners explicitly aware of the wrist strategy's operating principle. Keywords: skill acquisition, balance, postural control, declarative knowledge, model observation. INTRODUCTION "You make a hollow back" or "open your shoulder angle" are regular phrases gymnastics coaches use to comment on defective handstand performances. Augmented feedback and instructions have obtained widespread acceptance to benefit motor skill learning (Schmidt & Lee, 2011). However, are suchlike explicit advice sufficient? And what is considered as crucial for learning handstands at all? With the handstand on the floor being an essential postural motor task in gymnastics Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 (Hedbavny, Sklenarikova, Hupka, & Kalichova, 2013), contemporary educational concepts on motor learning in handstand acquisition are suggested to respond these questions from interdisciplinary perspectives. In general, a biomechanical point of view promises valuable knowledge regarding essential qualities for successful handstand performances. Due to modern technology allowing mobile data collection during sport-specific movements (Vogt, Kato, Schneider, Türk, & Kanosue, 2017) and handstand performances in particular (Blenkinsop, Pain, & Hiley, 2016), decoding motor behavior in handstands has still received considerable attention in recent research (Blenkinsop, Pain, & Hiley, 2017; Kochanowicz et al., 2017; Kochanowicz et al., 2018). Based upon previous studies addressing upright and inverted stance dynamics (Kerwin & Trewartha, 2001; Yeadon & Trewartha, 2003), it is meanwhile well accepted that postural control mechanisms during handstand on a plane surface initially depend on contributing torques by the most inferior joint (i.e., wrist). Compared to other involved muscles, wrist flexor torques have recently been reported to show the highest mean EMG activity (i.e., 60% RMS normalized to an isometric MVC) in handstand balances (Kochanowicz et al., 2018). In case of increasing postural oscillations, synergistic shoulder and hip torques are additionally used to maintain balance control (Hedbavny et al., 2013). With this, Gautier, Marin, Leroy and Thouvarecq (2009) accentuated that angular hip joint adjustments are only used by low-level gymnasts, whereas handstand performances executed by high-level experts are characterized by corrective movements in the most inferior joints. However, regardless of the applied balance control strategy, movement patterns facilitating maintenance in handstand benefit from visual control (Asseman & Gahery, 2005; Gautier, Thouvarecq, & Chollet, 2007) and remain the center of mass (CoM) vertically above the base of support (Kerwin & Trewartha, 2001). Thus, skill-related motor behavior in handstand can be modelled as a singlesegment inverted pendulum (Blenkinsop et al., 2017) with the body balanced above the wrists as one steady object (Yeadon & Trewartha, 2003). This in mind, high-level handstand performances and, thus, practice are suggested to approach two fundamental motor skills; (1) postural modulations leading to a linear rigid body configuration, (2) balance control abilities due to wrist flexor muscular activation. Considering that several biomechanical and motor behavioral studies only deal with handstand performances of experienced gymnasts (e.g., Gautier et al., 2009), current research on handstand skill acquisition in unexperienced learners indicates that the importance of these two respective aspects has rather been neglected in literature. Instead, several reports focused on the know-how of handstand coaching, referring to the expedient application of general knowledge about augmented feedback (e.g., Post, Aiken, Laughlin, & Fairbrother, 2016; Veit, Jeraj, & Lobinger, 2016) and observational learning (e.g., Andrieux & Proteau, 2016; Blandin, Lhuisset, & Proteau, 1999; Braun, 2016; Breslin, Hodges, & Williams, 2009; Buchanan & Dean, 2014; Hayes, Hodges, Scott, Horn, & Williams, 2006; Janelle, Champenoy, Coombes, & Mousseau, 2003; Laguna, 2008; Rohbanfard & Proteau, 2011). Regarding handstand acquisition, there are some studies investigating verbal (Masser, 1993), tactile (Croix, Lejeune, Anderson, & Thouvarecq, 2010) and several combined feedback (e.g., tactile-verbal and visual-comparative feedback; Rohleder & Vogt, 2018b) and observational learning strategies (Maleki, Nia, Zarghami, & Neisi, 2010). For example, Croix et al. (2010) suggest light finger contact on the thigh to increase gymnasts' balance Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 abilities in handstands. Previously, Masser (1993) observed enhanced handstand performances in students evoked by the verbal instruction "shoulder over your knuckles". With these studies providing substantial insights into enhanced handstand education, research dealing with the weighting of crucial training contents (i.e. wrist work, postural adaptations) with respect to explicit (EL) and implicit learning (IL; Sun, Merrill, & Peterson, 2001) is still lacking. Aiming for efficient coaching, it is well accepted that the complexity of potential effects regarding EL and IL has to be taken into account, particularly with respect to adversely affected performances due to reinvestment (Lam, Maxwell, & Masters, 2009; Malhotra, Poolton, Wilson, & Omuro, 2015; Masters & Maxwell, 2008; Verburgh, Scherder, van Lange, & Oosterlaan, 2016). While reports by Uzunov (2008) as well as Rohleder and Vogt (2018) suggest that wrist work and a proper body line seem to be mutually dependent for successful handstands, current training methodology privileges the explicit development of the postural alignment preceding wrist-related practice (Uzunov, 2008). Considering recent reports suggesting the wrist strategy to be the most dominant control strategy even in perturbed handstand balances (e.g., Blenkinsop et al., 2017), the relatively low status of explicit hand balance abilities in current handstand educational concepts may be challenged. Conscious of facilitated balance practicing due to preceding learning of postural stabilization, however, investigations on adapted EL-based handstand coaching providing declarative knowledge regarding wrist usage, accompanied by only implicit postural training, remain to be elucidated in consideration of different skill levels (Gautier et al., 2009; Kochanowicz et al., 2018; Vogt et al., 2017). Approaching altered learning of skill-related motor tasks compared to predominant handstand training, this study aimed to examine motor learning effects of explicit wrist strategy coaching on handstand performances in skilled and less skilled novices. With respect to presumably existing non-declarative knowledge regarding the wrist strategy in skilled novices, it is hypothesized that (1) less skilled compared to skilled novices show training-induced increases in balance time that are related to enhanced postural control patterns. (2) Less skilled compared to skilled novices are further hypothesized to show training-induced enhancements in postural performances that reflect beneficial effects of IL regarding an aligned body configuration. METHODS Thirty-two volunteering sport students (17 females: Mage = 20.71, SD = 0.99 years; Mheight = 167.85, SD = 7.38 cm; Mweight = 59.68, SD = 7.17 kg; 15 males: Mage = 22.07, SD = 1.67 years; Mheight = 183.40, SD = 5.77 cm; Mweight = 77.93, SD = 7.27 kg) were recruited from university courses to participate in this study. Volunteers who gained gymnastics experience well beyond school classes (i.e., particular history in organized gymnastics training systems) were excluded a priori. However, passing the university's physical aptitude test guaranteed participants' fundamental skill-related experience regarding the lunge entry and swing up to handstand movement. The study received approval by the University's Human Research Ethics committee in compliance with the Declaration of Helsinki. All participants provided written consent. Unaware of the experimental hypotheses, participants completed two experimental protocols (pre- and post-test) each comprising three trials (Gautier et al., 2007) of swinging up to handstand on the floor. Withholding augmented advice addressing a resistant linear body configuration, each participant was instructed beforehand to aim for long Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 maintenance in handstand with their feet kept close together. Following individual warm up and preparation (handstands were not permitted), one single handstand rehearsal served as familiarization preceding test trials (Rohleder & Vogt, 2018a). In absence of any rules describing a standardized termination of the handstand position, participants were allowed to roll over, to leave the handstand sideways or to place their feet back. Any changes of the hand position during handstand led to discontinuation of the trial. Approaching sport-specific conditions in consideration of a user-orientated environment (Rohleder & Vogt, 2018b), the experiment was conducted in the University's gym using a certified gymnastics mat to perform the handstand trials. A rectangular corridor (dimensions: 80 x 30 cm) was affixed on the mat using white tape to standardize the position of hand support during handstand (Figure 1C). Pre-tests were performed as described within one week (i.e., week 1; Figure 1A). Subsequently, participants received a video-tutorial providing concise declarative knowledge concerning the operating principle of the wrist strategy for successful handstand balances (Blenkinsop et al., 2017; Kerwin & Trewartha, 2001; Yeadon & Trewartha, 2003). Taking the current stage of research concerning effective observational learning into account, the tutorial used a mixed-model approach (Rohbanfard & Proteau, 2011) presenting one failed and one successful wrist-controlled hand balance. Video sequences were accompanied by screenshots and verbal comments (Janelle et al., 2003) raising participants' awareness for wrist-related knowledge of performance (Laguna, 2008) and models' skill level (Andrieux & Proteau, 2016). Limiting attentional cueing to wrist-related features ensured explicit and consistent focusing on the wrists' skill-related relevance (Breslin et al., 2009; Buchanan & Dean, 2014). Hence, advice stressing critical cues to facilitate postural adjustments were not taught. Combining observational learning with physical practice (Blandin et al., 1999), participants completed five training sessions within the following three weeks (Maleki et al., 2010; Masser, 1993) including two sets of six different exercises. Based on afore conveyed expertise, all exercises intended skill-related motor control in terms of wrist flexor activation during handstand on a plane surface. Partially inspired by Uzunov (2008), exercises (E1-E6) were designed as follows (Figure 1B): • E1: Winding up a rope (due to wrist flexions) which is connected with a weight plate (1.25 kg). • E2: Oblique standing with straight arms parallel next to the ears and repeated hand pushing against a wall leading to sole wall touch by the fingertips. • E3: Knee rest and repeated palmar flexion of the wrists (forehand view) with a weight plate (1.25 kg) in each hand. • E4: Free practicing of the lunge entry and swing up to handstand aiming for a vertically placed CoM above the fingers. Self-controlled video feedback (delay: 12 sec) was provided using a tablet PC (iPad Air) (Post et al., 2016). • E5: Free standing with straight arms parallel next to the ears and performing repeated palmar flexions of the wrists with a weight plate (1.25 kg) in each hand. • E6: Lunge entry and swing up to handstand leaning the thighs against a bar (fingertips approximately 10 cm away from the bar's perpendicular) and removal of the thighs from bar contact (Croix et al., 2010) due to sole wrist flexions. Pushing the bar actively by the legs was explicitly prohibited. In accordance with the video-tutorial, exercises aimed for explicit wrist flexor activation triggered by locating the CoM vertically above the support surface (i.e., E4 and E6). In favor of visual control (Asseman & Gahery, 2005), participants Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 were encouraged to gaze their hands throughout all exercises. Referring to recommended training loads (Uzunov, 2008), each exercise was executed for approximately thirty seconds followed by a self-determined recovery period (< 30 s). Postural adaptations and a controlled dosing (i.e., self-paced velocity) of the swing up to handstand movement were only addressed implicitly without any explicit cueing provided by the experimenters. After completion of the third training week, the post-test was conducted during the following week in conformity with the pre-test's modalities (Figure 1A). Figure 1. Schematic view displaying (A) the experimental design of the overall procedures, (B) the exercises performed within the training interventions (here: accurate execution quality by an experienced gymnast to emphasize the basic intention of the exercises best possible) and (C) the experimental setup for data collection. Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 Following a practical approach, a tablet PC (Apple iPad Air, frequency: 60 Hz; resolution: 1080p) was used to record all trials capturing the movement task in the sagittal plane. The tablet PC was fixed to a tripod at a height of one meter (distance to the middle of support surface: 4.85 m) ensuring a standardized point of view (Rohleder & Vogt, 2018a; Figure 1C). Tape markers were attached to the following anatomical landmarks to track the position of crucial body segments; 1: wrist (ulnar-styloid process), 2: shoulder (posterior deltoid), 3: hip (femur greater trochanter), 4: knee (lateral epicondyle of femur), 5: ankle (fibula apex of lateral malleolus). Based on recorded video sequences, kinematic data were further determined using the Kinovea 0.8.15 software. In line with official criteria of the International Gymnastics Federation (FIG) concerning point deductions for handstand performances due to angular deviations (FIG, 2017), the balance time in handstand was measured using a corridor limited by a deviation of 15° to each side of vertical line above the wrists (Rohleder & Vogt, 2018a). With this, time measuring was started when both legs (ankles and knees) initially entered the corridor following the swing up to handstand. Moreover, time measuring was discontinued under the following conditions; 1: corridor exit of the feet or knees, 2: Initiation of the rolling over movement (i.e., incipient arm bending or rolling in of the head), 3: Initiation of the back placing of the feet (i.e., incipient increase of the maximum reduced feet gap). The quality of postural execution was evaluated by four independent artistic gymnastics experts (i.e., two nationally licenced coaches and two judges; two male and female each) assigning scores between 0.0 and 10.0 points to each trial matching current FIG rules (FIG, 2017). For this, experts were briefed to completely neglect participants' balance time in handstand. Instead, experts were instructed to particularly evaluate unsteadiness, slightly bent arms and legs (in accordance with the criteria described in 2.4.1) as well as angular deviations from the perfect hold position (FIG, 2017). Trials of each participant (i.e., 3x pre-test; 3x post-test) were presented to the experts in randomized order. Additionally, experts were asked to assign one of five balance control strategies to each trial with respect to the externally visible postural correction mechanisms in anterior-posterior direction. Balance control strategies were characterized as follows (Figure 2): • Wrist strategy (4 points): robust body configuration from wrists to ankles with apparent wrist-controlled balance corrections • Shoulder strategy (3 points): robust body configuration from shoulders to ankles with apparent shoulder-controlled balance corrections • Shoulder-hip-coupling (2 points): apparent contrary shoulder- and hip-controlled balance corrections • Hip strategy (1 point) robust body configuration from wrists to the hip with apparent hip-controlled balance corrections • "No strategy" (0 points): no apparent motor skills to maintain the handstand position Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 Figure 2. Categories of anterior-posterior balance control strategies in handstand assigned by the experts (inspired by Blenkinsop et al., 2017 and Gautier et al., 2009). Classification: The larger the distance of the primarily correcting joint to the support surface, the lower the points for the applied balance control strategy \ 1 } { ? 4 points 3 points 2 points 1 point 0 points Goniometric analyses focused on large body joints (i.e., shoulder and hip joint) which, except for the wrists, have been suggested to primarily contribute to postural control during handstand performances (Gautier et al., 2009). To determine joint-related angular changes during handstand, the individual optimum of each handstand trial was identified (Masser, 1993) based on predefined criteria derived from Rohleder & Vogt (2018b); (1) in case of rolling over trials, the moment where the participants' feet reached their highest point was set as optimum, (2) in case of trials characterized by side or back placement of the feet, the moment of incipient leg straddling or spreading served as optimum. Due to conflicting schedules or unpredicted injuries caused by their sport studies, six participants did not complete the full experimental procedure and were excluded from statistical analyses. After checking interval scaled variables (balance time, shoulder angle, hip angle) for outliers using the 2o-method, remaining participants (n = 25) were assigned to two groups based on rankings of attained pretest mean balance times. Inspired by Vogt et al. (2017), the ranking list was divided into a long-balancing half (Mbalance time = 1.13 s, SD = .42, n = 13) and a short-balancing half (Mbalance time = .41 s, SD = .18, n = 12). According to this bisection of rankings (unpaired t-test revealed t[23] = 5.40, p < .01), participants were classified as skilled (long-balancing half) and less skilled novices (short-balancing half) for further analyses. Repeated measures analyses of variance (ANOVAs) were computed for each dependent interval scaled variable (i.e., balance time, shoulder and hip angle) including group (skilled vs. less skilled) as between factor and test (pre- vs. post-test) as within factor. Further, pairwise comparisons (i.e., paired and unpaired t-tests) were calculated post hoc. For ordinal scaled variables (i.e., postural execution, balance control strategy), Wilcoxon signed-ranks tests and Mann-Whitney-U-tests were immediately performed as pairwise comparisons. All calculations for pairwise comparisons were adjusted for multiple testing by applying Bonferroni-Holm corrections. Kendall's coefficient of concordance (W) was calculated to ensure interrater reliability of the ordinal scaled variables concerning expert judgements (postural execution: W(149) = .787, p < .01; Balance control strategy: W(149) = .487, p < .01 ). Partial eta-squared (n2p) was used to identify ANOVAs' effects, whereas Cohen's d effect sizes were calculated to interpret pre-to-post changes. Statistical analyses were performed using Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 the SPSS 25.0 software (International Business Machines Corp., Armonk, NY, USA). The level of significance was set at p < .05. Data in the text, tables and figures are presented as means (M) ± standard deviations (SD). RESULTS Balance time Repeated measures ANOVA revealed significant interactions between factors group and test, F(\, 23) = 6.40, p < .05, n2p = .22; however, main effects revealed no differences for tests, F(\, 23) = 1.88, p > .05, n2p = 08. Post hoc tests showed significant enhanced balance times for less skilled, t(11) = 3.93, p < .01, d = 1.30, but not for skilled novices, ¿(12) = .69, p > .05, d = .27; further, post hoc tests showed significant group differences in the pretest, <23) = 5.40, p < .01, which were not obtained for the post-test, t(23) = .92, p > .05 (Figure 3; Table 1). Postural execution Wilcoxon signed-ranks test showed significantly increased scores for postural execution in less skilled novices, Z = -2.67, p < .05, d = .79, but not for skilled novices, Z = -1.71, p > .05, d = .47; further, Mann- Whitney-U-test showed significant group differences in the pre-test, U = 22.50, p < .01, which were not obtained for the posttest, U = 77.50,p > .05 (Figure 3; Table 1). Balance control strategy Wilcoxon signed-ranks test showed significantly increased scores for balance control strategy in less skilled novices, Z = -2.43, p < .05, d = .91, but not for skilled novices, Z = -1.15, p > .05, d = .43; further, Mann-Whitney-U-test showed no significant group differences in the pretest, U = 45.50, p > .05, and in the posttest, U = 63.00,p > .05 (Figure 3; Table 1). Goniometric analysis Shoulder angle: repeated measures ANOVA revealed no interactions between factors group and test, F(1, 23) = 1.43, p > .05, n2p = .06; further, main effects revealed no differences for tests, F(1, 23) = 1.86, p > .05, n2p = .08 (Figure 3; Table 1). Hip angle: repeated measures ANOVA revealed no significant interactions between factors group and test, F(1, 23) = 5.09, p > .05, n2p = .18; further, main effects revealed no differences for tests, F(1, 23) = .01, p > .05, n2p = .00. (Figure 3; Table 1). Figure 3. Interaction plots displaying means ± standard deviations for group-related changes (less skilled vs. skilled novices) of factors balance time, postural execution, postural control strategy as well as shoulder and hip joint angle. Level of significance in pairwise comparisons (adjusted to Bonferroni-Holm-corrections): *p < .05. Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 Table 1 Changes in handstand performances (means ± standard deviations) during pre- and post-test for less skilled and skilled novices. Less skilled Skilled Pre-test Post-test Pre-test Post-test Balance time [s] 0.41 ± 0.18 Postural execution [points] 4.13 ± 1.23 Balance control strategy [points] 1.65 ± 0.53 Shoulder angle [°] 164.47 ± 8.40 Hip angle [°]_180.45 ± 16.45 0.83 ± 0.42 5.18 ± 1.42 2.22 ± 0.72 167.58 ± 7.83 185.86 ± 15.39 1.13 ± 0.42 5.74 ± 1.21 2.23 ± 0.71 166.59 ± 4.95 189.69 ± 8.48 1.00 ± 0.48 5.19 ± 1.12 1.94 ± 0.61 166.79 ± 6.59 184.74 ± 11.44 DISCUSSION The present study aimed to examine motor learning effects of explicit wrist strategy coaching on handstand performances in skilled and less skilled novices. With pre-test mean balance time rankings serving as criterion for skill level-based group assignment (i.e., skilled vs. less skilled novices), enhanced mean balance times, scores for postural execution and postural control strategies were observed in less skilled compared to unaffected parameters in skilled novices. Furthermore, in both groups changes in shoulder and hip joint angles failed significance. According to our initial first hypothesis (1), increased mean balance times in less skilled novices reflect beneficial effects of EL evoking declarative knowledge (Sun et al., 2001) in entirely unexperienced learners regarding the wrist strategies' operating principle (Blenkinsop et al., 2017; Kerwin & Trewartha, 2001; Yeadon & Trewartha, 2003). With this, increased handstand balance times in less-skilled novices support combined physical practice with observational learning and augmented feedback to accelerate general (Magill, 2014; Schmidt & Lee, 2011), gymnastics-related (Braun, 2016; Veit et al., 2016) and even handstand-specific skill development (Maleki et al., 2010). However, while all participants (i.e., skilled and less skilled) received the same literature-based videotutorial (e.g., Andrieux & Proteau, 2016; Science of Gymnastics Journal Janelle et al., 2003; Rohbanfard & Proteau, 2011) and training intervention (e.g., Asseman & Gahery, 2005; Post et al., 2016; Uzunov, 2008), unaffected mean balance times in skilled compared to less skilled novices have to be discussed in light of large standard deviations and EL-induced performance-influencing factors (Lam et al., 2009; Malhotra et al., 2015). Taking recommended and, thus, implemented consistency in attentional directing and movement task usage into consideration (Breslin et al., 2009; Buchanan & Dean, 2014), there are reasons to assume that the present study's findings may be attributed to the reinvestment theory by Masters and Maxwell (2008) suggesting declarative task-relevant knowledge (i.e., wrist strategy usage) to interfere already automated motor processes which were presumably present in the skilled group prior to the intervention. Facing this, it seems reasonable that task-relevant knowledge may occasionally impair more skilled handstand balances (Masters & Maxwell, 2008), apparently independent of the approached coaching focus regarding skill-related motor tasks (e.g., wrist strategy usage vs. postural stabilization; Rohleder & Vogt, 2018a). Nevertheless, although showing negative tendencies, it has to be taken into account that impaired balance times and balance control strategies in skilled novices failed significance. Thus, interpretations of 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 present effects with respect to reinvestment can only serve as assumptions and need further research. Furthermore, with Kochanowicz et al. (2018) and Gautier et al. (2009) reporting skill level dependencies of balance control, groups are considered to reflect emerging differences regarding altered motor behavior as a result of the training intervention. However, with respect to this, present findings disclose inconsistencies in view of comparable studies. For example, facilitated handstand balances due to light thigh touch during testing were revealed for experienced gymnasts (Croix et al., 2010). On the one hand, this contradicts advantageous balance performances in less skilled compared to skilled participants in the present study following training including a tactile advice on the thigh (E6). On the other hand, in the present study contact on the thigh was only applied during training, but not during testing which interferes comparability. Furthermore, Croix et al. (2010) then again applied a lateral touch compared to a dorsal touch in the present study and related enhanced balancing to the lateral, but not to the anterior-posterior direction which is essentially approached in the present study. With this, skill level-related comparisons must take the level of experience and setup-alterations into account in more detail. Using laboratory setups, Kochanowicz et al. (2018) compared two experienced groups (young and adult gymnasts) as well as Gautier et al. (2009) comparing high-level and low-level experts who were all able to maintain in the handstand for at least twenty seconds. Thus, although filling a stated research gap according to efficient coaching on wrist usage in handstand acquisition (Rohleder & Vogt, 2018a), broad evidence regarding motor behavior in experienced gymnasts (Blenkinsop et al., 2017; Gautier et al., 2007; Gautier et al., 2009; Kochanowicz et al., 2017; Kochanowicz et al., 2018) remains indistinct for unexperienced performers and, thus, complicates integrating our findings into an appropriate context of literature. This encourages to further elucidate novices motor behavior in response to altered educational concepts approaching EL and IL effects in handstand acquisition. According to our second hypothesis (2), increased postural execution in less skilled novices suggests explicit wrist strategy coaching to even induce posture-related IL leading to more aligned body configurations. These findings confirm well-accepted knowledge regarding IL benefits (Lam et al., 2009; Verburgh et al., 2016) and are additionally in line with comparable studies reporting enhanced postural performances in handstands following feedback- and observation-based interventions (Maleki et al., 2010; Masser, 1993; Rohleder & Vogt, 2018b). However, while the present study revealed enhanced postural configurations following IL, Maleki et al. (2010) related observational learning benefits to additional (explicit) verbal teaching. In addition, the verbal cue „shoulder over your knuckles" used by Masser (1993) provides declarative knowledge in terms of postural adaptations which is in contrary to the present study's approach revealing implicit body alignment in less skilled novices. Moreover, Maleki et al. (2010) assumed the skill level to influence observational and physical practice benefits, which was partially reflected by unaffected postural execution in skilled novices. Additionally, from a goniometric perspective, shoulder joint angles remained unaffected in both groups which is in conflict to a previous study by Rohleder and Vogt (2018b) reporting feedback-induced shoulder angle changes in handstand positions in the absence of hip-related effects. Challenging these contradictory reports, it seems reasonable that the studies' divergent primary objectives which were explicitly taught to the participants a priori may have caused different postural adaptations. While Rohleder and Vogt (2018b) Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 explicitly addressed high postural execution quality neglecting maintenance in handstands, the participants in the present study were exclusively instructed to aim for a long balance time, which may eventually induce different movement patterns regarding postural control. Additionally, although failing significance, slightly increased hip angles in less skilled novices provide reasons to assume that unexperienced learners implicitly tend to use opening the hip joint in order to locate the CoM vertically above the support surface. Although the comparability of skill-levels is partially limited, this assumption is in line with Gautier et al. (2009) reporting low-level experts to prefer increased hip joint involvement to coordinate postural control. Necessitating further research, this might indicate implicit postural adaptations in response to EL regarding wrist usage. In summary, the present study's findings legitimize the underlying approach of handstand coaching concepts strictly geared to skill-related motor behavior in terms of an increased explicit focus on wrist usage. Confirming initial assumptions from previous reports (Rohleder & Vogt, 2018a), explicit wrist strategy coaching neglecting any kind of posture-related advice benefits comprehensive handstand acquisition in less skilled novices including enhanced balance and, remarkably, enhanced quality in handstands' postural execution. This is, at least in parts, in contrary to Uzunov (2008) suggesting rigid body line development preceding balance training. However, efficacy of wrist-related handstand coaching seems to be ineffective to relatively skilled performers. While these negative effects may again be discussed with respect to EL effects on reinvestment (Masters & Maxwell, 2008), there is a further need to clarify if the practical training or the taught "know how" in the video-tutorial (or, presumably, the combination of both) essentially caused observed changes in handstand performances. Although studies prefer combined observational learning and physical practice (Blandin et al., 1999; Hayes et al., 2006), there are few studies indicating positive EL effects even without practical training (Maleki et al., 2010; Rohleder & Vogt, 2018b). This, for example, suggests further research to exclusively address observational learning introducing novices to the wrist strategy of handstand balances. LIMITATIONS Considering that the small number of participants generally limits the power of the present study, the recruited sample size seemed to be appropriate in relation to comparable studies in this research field (e.g., Croix et al., 2010). Nevertheless, we are well aware that a missing group of participants receiving no kind of coaching and practice represents a methodological limitation. Although further investigations are suggested to take an additional control group into consideration, the premeditated group assignment was deemed appropriate with respect to our comparative approach on motor learning effects in dependence on novices' skill level. Furthermore, although the unsteadiness of durations in handstand position (leading to high standard deviations) limits the present approach, we see this limitation as a necessary expense for a study aiming to serve as a kick off for applied research on the coaching of unexperienced handstand learners. Additionally, choosing a practice-orientated setup may be discussed in relation to reliable data collection. However, in view of comparable laboratory-based studies with similar shortcomings (e.g., Gautier et al., 2007; comparable low-frequency video sampling), reserves in data reliability are disproportionate to validity-related benefits of the chosen setup providing familiar and safe conditions to allow natural movement execution which is uninfluenced by impeding laboratory framework. Finally, it Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 has to be taken into account that the differentiation between judging postural execution and balance control strategies was a necessary, but difficult task for judges which probably further limits the present study. CONCLUSIONS To conclude, the present study investigated effects of explicit wrist strategy coaching on novices' handstand performances depending on the skill level. Present findings (i.e., increased balance time and execution quality) suggest practitioners to make entirely unexperienced learners explicitly aware of the wrist strategy's operating principle. This is with respect to a no less important coaching of postural stabilization. With this study extending practice-oriented knowledge of efficient motor learning in handstand acquisition, appropriate educational strategies need further investigations approaching insights into skill-related motor behaviour, especially in unexperienced handstand learners. ACKNOWLEDGEMENTS The authors would like to express their appreciation to all participants for their enthusiastic responses and reliable efforts in favor of the present study. This study was supported by an international grant of the German Academic Exchange Service (DAAD project no. 57320531) awarded to the supervising author. REFERENCES Andrieux, M., & Proteau, L. (2016). Observational Learning: Tell Beginners What They Are about to Watch and They Will Learn Better. Frontiers in Psychology, 7, 51. doi:10.3389/fpsyg.2016.00051 Asseman, F., & Gahery, Y. (2005). Effect of head position and visual condition on balance control in inverted stance. Neuroscience Letters, 375(2), 134137. doi:10.1016/j.neulet.2004.10.085 Blandin, Y., Lhuisset, L., & Proteau, L. (1999). Cognitive Processes Underlying Observational Learning of Motor Skills. The Quarterly Journal of Experimental Psychology, 52(4), 957-979. doi:10.1080/027249899390882 Blenkinsop, G. M., Pain, M. T. G., & Hiley, M. J. (2016). Evaluating feedback time delay during perturbed and unperturbed balance in handstand. Human Movement Science, 48, 112-120. doi:10.1016/j.humov.2016.04.011 Blenkinsop, G. M., Pain, M. T. G., & Hiley, M. J. (2017). Balance control strategies during perturbed and unperturbed balance in standing and handstand. Royal Society Open Science, 4(7), 161018. doi:10.1098/rsos.161018 Braun, C. (2016). Observational Learning in the Context of Skill Acquisition. In T. Heinen, I. Čuk, R. Goebel, & K. Velentzas (Eds.), Gymnastics Performance and Motor Learning: Principles and Applications (pp. 34-43). New York: Nova Science Publsihers, Inc. Breslin, G., Hodges, N. J., & Williams, A. M. (2009). Effect of information load and time on observational learning. Research Quarterly for Exercise and Sport, 80(3), 480-490. doi:10.1080/02701367.2009.10599586 Buchanan, J. J., & Dean, N. (2014). Consistently modeling the same movement strategy is more important than model skill level in observational learning contexts. Acta Psychologica, 146, 19-27. doi:10.1016/j.actpsy.2013.11.008 Cicchetti, D. V. (1994). Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychological Assessment, 6(4), 284-290. doi:10.1037/1040-3590.6.4.284 Croix, G., Lejeune, L., Anderson, D. I., & Thouvarecq, R. (2010). Light fingertip contact on thigh facilitates handstand balance in gymnasts. Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 Psychology of Sport and Exercise, 11(4), 330-333. doi:10.1016/j.psychsport.2010.02.003 FIG (2017). 2017 Code of Points: Men's Artistic Gymnastics. Lausanne: Fédération Internationale de Gymnastique. Gautier, G., Thouvarecq, R., & Chollet, D. (2007). Visual and postural control of an arbitrary posture: The handstand. Journal of Sports Sciences, 25(11), 1271-1278. doi:10.1080/02640410601049144 Gautier, G., Marin, L., Leroy, D., & Thouvarecq, R. (2009). Dynamics of expertise level: Coordination in handstand. Human Movement Science, 28(1), 129140. doi:10.1016/j.humov.2008.05.003 Hayes, S. J., Hodges, N. J., Scott, M. A., Horn, R. R., & Williams, A. M. (2006). Scaling a motor skill through observation and practice. Journal of Motor Behavior, 38(5), 357-366. doi:10.3200/JMBR.38.5.357-366 Hedbâvny, P., Sklenarikovâ, J., Hupka, D., & Kalichovâ, M. (2013). Balancing the handstand on the floor. Science of Gymnastics Journal, 5(3), 6980. Janelle, C. M., Champenoy, J. D., Coombes, S. A., & Mousseau, M. B. (2003). Mechanisms of attentional cueing during observational learning to facilitate motor skill acquisition. Journal of Sports Sciences, 21(10), 825-838. doi:10.1080/0264041031000140310 Kerwin, D. G., & Trewartha, G. (2001). Strategies for maintaining a handstand in the anterior-posterior direction. Medicine and Science in Sports and Exercise, 33(7), 1182-1188. doi:10.1097/00005768-200107000-00016 Kochanowicz, A., Niespodzinski, B., Marina, M., Mieszkowski, J., Biskup, L., & Kochanowicz, K. (2018). Relationship between postural control and muscle activity during a handstand in young and adult gymnasts. Human Movement Science, 58, 195-204. doi:10.1016/j.humov.2018.02.007 Kochanowicz, A., Niespodzinski, B., Marina, M., Mieszkowski, J., Kochanowicz, K., & Zasada, M. (2017). Changes in the Muscle Activity of Gymnasts During a Handstand on Various Apparatus. Journal of Strength and Conditioning Research. Advance online publication. doi:10.1519/JSC.0000000000002124 Laguna, P. L. (2008). Task complexity and sources of task-related information during the observational learning process. Journal of Sports Sciences, 26(10), 1097-1113. doi:10.1080/02640410801956569 Lam, W. K., Maxwell, J. P., & Masters, R. (2009). Analogy Learning and the Performance of Motor Skills under Pressure. Journal of Sport and Exercise Psychology, 31(3), 337-357. https://doi.org/10.1123/jsep.3L3.337 Magill, R. A. (2014). Motor learning and control: Concepts and Applications (10th ed.). Boston: McGraw-Hill. Maleki, F., Nia, P., Zarghami, M., & Neisi, A. (2010). The Comparison of Different Types of Observational Training on Motor Learning of Gymnastic Handstand. Journal of Human Kinetics, 26, 13-19. doi:10.2478/v10078-010-0043-0 Malhotra, N., Poolton, J. M., Wilson, M. R., & Omuro, S. (2015). Dimensions of Movement-Specific Reinvestment in Practice of a Golf Putting Task. Psychology of Sport and Exercise, 18, 1-8. doi:10.1016/j.psychsport.2014.11.008 Masser, L. S. (1993). The Comparison of Different Types of Observational Training on Motor Learning of Gymnastic Handstand. Journal of Teaching in Physical Education, 12, 301-312. Masters, R., & Maxwell, J. (2008). The theory of reinvestment. International Review of Sport and Exercise Psychology, 1(2), 160-183. doi:10.1080/17509840802287218 Post, P. G., Aiken, C. A., Laughlin, D. D., & Fairbrother, J. T. (2016). Self-control over combined video feedback and modeling facilitates motor learning. Science of Gymnastics Journal 209 Science of Gymnastics Journal Rohleder J., Vogt T.: EFFICACY OF WRIST STRATEGY COACHING ON HANDSTAND. Vol. 11 Issue 2: 209 - 222 Human Movement Science, 47, 49-59. doi:10.1016/j.humov.2016.01.014 Rohbanfard, H., & Proteau, L. (2011). Learning through observation: A combination of expert and novice models favors learning. Experimental Brain Research, 215(3-4), 183-197. doi:10.1007/s00221-011-2882-x Rohleder, J., & Vogt, T. (2018a). Performance control in handstands: Challenging entrenched coaching strategies for young gymnasts. International Journal of Performance Analysis in Sport, 4(10), 1-15. doi:10.1080/24748668.2018.1440459 Rohleder, J., & Vogt, T. (2018b). Teaching novices the handstand: A practical approach of different sport-specific feedback concepts on movement learning. Science of Gymnastics Journal, 10(1), 29-42. Schmidt, R. A., & Lee, T. D. (2011). Motor Control and Learning: A Behavioral Emphasis (5th ed.). Champaign, IL: Human Kinetics. Sun, R., Merrill, E., & Peterson, T. (2001). From implicit skills to explicit knowledge: a bottom-up model of skill learning. Cognitive Sciences, 25, 203-244. doi:10.1016/S0364-0213(01)00035-0 Uzunov, V. (2008). The handstand: A four stage training model. Gym Coach, 2, 52-59. Veit, J., Jeraj, D., & Lobinger, B. H. (2016). Augmented Feedback for Movement Error Correction in Gymnastics. In T. Heinen, I. Čuk, R. Goebel, & K. Velentzas (Eds.), Gymnastics Performance and Motor Learning: Principles and Applications (pp. 45-52). New York: Nova Science Publsihers, Inc. Verburgh, L., Scherder, E. J. A., van Lange, P. A. M., & Oosterlaan, J. (2016). The key to success in elite athletes? Explicit and implicit motor learning in youth elite and non-elite soccer players. Journal of Sports Sciences, 34(18), 17821790. doi:10.1080/02640414.2015.1137344 Vogt, T., Kato, K., Schneider, S., Türk, S., & Kanosue, K. (2017). Central neuronal motor behaviour in skilled and less skilled novices - Approaching sports-specific movement techniques. Human Movement Science, 52, 151-159. doi:10.1016/j.humov.2017.02.003 Yeadon, M. R., & Trewartha, G. (2003). Control Strategy for a Hand Balance. Motor Control, 7(4), 421-442. doi:10.1123/mcj.7.4.421 Corresponding author: Jonas Rohleder German Sport University Cologne Institute of Professional Sport Education and Sport Qualifications Am Sportpark Müngersdorf 6 50933 Cologne Germany Tel: +49 (0) 221 - 4982 4160 Fax: +49 (0) 221 - 4982 8261 E-mail: j.rohleder@dshs-koeln.de Science of Gymnastics Journal 209 Science of Gymnastics Journal Sands W.A. et al.: COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES. Vol. 11 Issue 2: 223 - 237 COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES OF TRAMPOLINES William A. Sands1, Madison K. Varmette1, Gregory C. Bogdanis2, Olyvia Donti2, Bryce V. Murphy3, Troy J. Taylor1 !U.S. Ski and Snowboard Association, Park City, USA 2School of Physical Education and Sport Science, National and Kapodistrian University of Athens, Athens, Greece 3U.S. Olympic Committee, Olympic Training Center, Colorado Springs, USA _Original article Abstract Trampoline use has skyrocketed in recent years in a variety of recreational contexts and among athletes in sports ranging from gymnastics and diving to skiing and snowboarding. The purpose of this study was to examine the bounce characteristics elicited by athletes bouncing on three types of trampolines. Tumbl Trak, Standard, and Super Tramp trampolines were assessed by 10 experienced trampoline and acrobatic athletes (5 males, 5 females). A triaxial accelerometer (250 Hz) characterized the 10 highest controlled bounces on each trampoline and each athlete. Repeated measures ANOVAs showed statistical differences in bounce characteristics: time from bounce start to peak acceleration (p<.001, q2 =0.82), time from peak acceleration to bounce end (p=.030, q2 =0.40), and total bounce time (p<0.001, q2 =0.78, jump height (p<.001, q2 =0.95) peak acceleration (p=.015, q2 =0.37), and flight time (p<.001, q2 =0.97). Average acceleration, force, and allometrically scaled average force were not statistically different (p>.140, q2=0.20). The stiffest trampoline with the least time values, peak accelerations, and jump heights was the Tumbl Trak, followed by the Standard trampoline, and Super Tramp, respectively. This information may help practitioners and others to understand the bounce behaviors of athletes on these types of trampolines. Keywords: trampoline, comparison, acceleration, jumping. INTRODUCTION Trampolines have been used in various forms from the early 1900s (Horne, 1968). The modern trampoline dates from 1936 and was developed by George Nissen and Larry Griswold (Ladue & Norman, 1954). The characteristics that make a trampoline are a flexible surface (i.e., bed) attached to dozens of springs which are in turn connected to a large round, square, or rectangular metal frame. The trampoline has undergone a number of designs that include different sizes, shapes, beds, and spring configurations for different purposes. Trampoline is an Olympic competitive event (International Gymnastics Federation, 2018). In spite of the relatively long history of trampolines, there is a paucity of scientific information about the mechanical behavior of the trampoline and the interactions of athletes with these apparatuses. Trampolines are used for recreation (Fisher, 2010), spatial orientation (Heinen, 2011), fitness exercise (Atterbom & MacLean, 1983; Cugusi et al., 2018), gymnastics (Heinen, 2011; Science of Gymnastics Journal 223 Science of Gymnastics Journal Sands W.A. et al.: COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES. Vol. 11 Issue 2: 223 - 237 Hondzinksi & Darling, 2001), diving (Kimball, 1999), medical treatment (Giagazoglou et al., 2013; Hahn, Shin, & Lee, 2015; Sahlberg & Strandvik, 2005), and as a competitive event (Esposito & Esposito, 2009; Jensen, Scott, Krustrup, & Mohr, 2013). A recent New York Times posting described a new activity called "Gtramp" which involves backyard and other trampoline use arising from skateboard, parkour, and youth social media activities (Kettler, 2018). Trampolines offer athletes the ability to rise as high as five or more meters in the air with minimal physical effort (Eager, Chapman, & Bondoc, 2012). The flight time of trampoline jumping enhances an athlete's ability to practice difficult skills, gain spatial awareness, and land on a soft trampoline bed. Trampoline beds are assumed to be soft and flexible. However, the "softness" of a trampoline bed represents a flawed understanding (Farquharson, 2012). The energy required to project the athlete high in the air is considerable and on descending and landing the energy from the flight should be absorbed and returned by the athlete's musculoskeletal system and the trampoline bed. Understanding how bouncing on trampolines may affect timing, acceleration, height, and energy exchange is largely unknown with the exception of some physics modeling (Blajer & Czaplicki, 2001; Chen et al., 2016; Yeadon & Hiley, 2017). A recent study of circus acrobats used a wearable three-dimensional accelerometer to measure accelerations in seven male acrobats during training and show performances. The results revealed that accelerations were statistically greater during training than shows. Moreover, accelerations were classified into categories of magnitude from approximately 1 g to more than 12 g (Barker, Burnstein, & Mercer, 2018). A conference presentation on trampoline measurements showed average peak accelerations of approximately 5 g (49 m/s2) and flight times ranging from 0.50 s to 0.54 s (Eager et al., 2012). The values from the Eager and colleagues (Eager et al., 2012) measurements are astonishingly low. The corresponding jump height for these values would be approximately 31 cm to 36 cm which is easily attainable in a vertical jump from the floor (Simons & Bradshaw, 2016b). A thesis investigating the relationship between trampoline bouncing and the countermovement vertical jump found no statistically significant relationship (Briggs, 2014). A study by the National Aeronautics and Space Administration (NASA) showed that trampoline bouncing by eight males at four heights with accelerometers on the ankle, forehead, and back resulted in accelerations of 3.0-7.0 g, 3.9-6.0 g, 3.05.6 g, respectively (Bhattacharya, McCutcheon, Shvartz, & Greenleaf, 1980). Trampolining has received considerable attention in terms of injury and injury prevention in Australia, Germany, New Zealand, and the U.S. (Ashby, Pointer, Eager, & Day, 2015; Chalmers, Hume, & Wilson, 1994; Hammer, Schwartzbach, & Paulev, 1981; Lewald, 1979; Sandler et al., 2011; Torg & Das, 1984). Moreover, studies of the benefits of trampolining include aerobic fitness, convenience, and balance (Atterbom & MacLean, 1983; Butler, 1969; Da Roza, Brandao, Mascarenhas, Jorge, & Duarte, 2015; Giagazoglou et al., 2013; Guillot & Collet, 2004; Hardy, Mullen, & Martin, 2001; Heitkamp, Horstmann, Mayer, Weller, & Dickhuth, 2001; Katch, Villanacci, & Sady, 1981; Ladue & Norman, 1954). However, analyses of bounce characteristics have seldom been addressed. In addition, there appear to be only a few studies comparing backyard trampolines, mini-trampolines, and full-size or competitive trampolines in terms of injury incidence and rates (Council On Sports & Fitness, 2012; Sands, Hondzinski, Shultz, & George, 1995; Torg & Das, 1985). It is our belief that the lack of information on the Science of Gymnastics Journal 223 Science of Gymnastics Journal Sands W.A. et al.: COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES. Vol. 11 Issue 2: 223 - 237 characteristics of bouncing on different trampolines has been ignored and merits research. The Center of Excellence facility of the U.S. Ski and Snowboard Association headquarters is unusual in that there are three different types of trampolines, expert trampolinists (i.e., national team Aerial, Moguls, and Half-Pipe Skiers along with Half-Pipe and Big-Air Snowboard), and expert coaching. Many of these athletes are former gymnasts and trampolinists. Moreover, these athletes regularly perform skills on trampolines that are equal or more difficult than competitive trampolinists. These athletes regularly perform quad-twisting triple somersaults (Aerials) and quadruple somersaults (Big-Air Snowboard). These skills are performed on unpredictable terrain, a variety of weather conditions, and landings on snow. Trampoline use has skyrocketed in recent years in a variety of recreational contexts and among athletes in sports ranging from gymnastics and diving to skiing and snowboarding (Ashby et al., 2015; Chalmers et al., 1994; Esposito & Esposito, 2009; Fisher, 2010). However, scientific understanding of the behavior of trampolines has not kept pace. Characterizing the interaction of trampoline-related activities and athletes may assist practitioners, scientists, and medical professionals in encouraging or discouraging use of trampoline bouncing for acrobatic athletes and others. The purpose of this study was to characterize the bounce behaviors elicited by three types of trampolines to determine the relative accelerations, durations, average forces and other factors. In spite of a relatively long history of trampoline use, and the fact that many catastrophic injuries occur in the center of the trampoline bed by highly-trained athletes (Torg, 1985; Torg & Das, 1984), it should be imperative to derive a more complete understanding of the workings of trampoline-athlete interactions while bouncing. We hypothesized that all of the trampoline types would show statistically different bounce characteristics. METHODS Participants: Ten experienced trampoline athletes from the U.S. Ski and Snowboard Aerials Team volunteered to participate in this study. The anthropometric information for the athletes was: five males (Mean ± SD; age 22.6 y ± 3.4 y; height 174.0 cm ± 5.0 cm; mass 73.2 kg ± 9.2 kg) and five females (Mean ± SD; age 19.8 y ± 2.8 y; height 160.2 cm ± 5.0 cm; mass 57.9 kg ± 4.8 kg). Equipment: Bounce accelerations provided by the athletes were obtained from three types of trampolines: Tumbl Trak (bed size = 1.52 m x 11.89 m, solid black bed, Tumbl Trak, Mount Pleasant, MI, USA), a standard competitive trampoline (bed size = 2.14 m x 4.27 m, two-string bed, Rebound Products, Thornhill, Ontario, Canada), Super Tramp (bed size = 3.05 m x 6.10 m, one-string bed, Rebound Products, Thornhill, Ontario, Canada). See Figures 1-3. Instrumentation: Accelerations were obtained from a PASCO Scientific triaxial accelerometer (PASCO Scientific, Roseville, CA, USA PS-3202) attached rigidly to a waist belt that was worn snugly about the waist of the athlete placing the accelerometer posterior to the lumbar spine at approximately the level of lumbar vertebrae L3 to L4 (Simons & Bradshaw, 2016a). Accelerometer placement has varied widely in experiments because of potential threats to stability from skin movement, subcutaneous fat, breathing, tissue inertia and many other factors (Simons & Bradshaw, 2016a). Placement of accelerometers on the upper back has been compared to the lower back among female participants in bilateral hopping and drop landings (Simons & Bradshaw, 2016a, 2016b) with better correlations with drop landings on a force platform arising from an upper back placement and better inter-day reliability arising from a low Science of Gymnastics Journal 223 Science of Gymnastics Journal Sands W.A. et al.: COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES. Vol. 11 Issue 2: 223 - 237 back placement (Simons & Bradshaw, 2016a, 2016b). Acceleration data were transmitted via Bluetooth to a laptop computer. Data were captured, displayed, and stored using the PASCO Capstone software (PASCO Scientific, Roseville, CA, USA, V1.11.1). The sampling rate was 250 Hz. Calibration was performed using gravitational vertical. Calibration was ensured by rotating the accelerometer systematically such that one of the three axes of the accelerometers was oriented to the line of gravity approximately 9.806 m/s2, while the remaining axes measured approximately 0 m/s2. Figure 1. Tumbl Trak trampoline. Figure 2. Standard trampoline. Procedures: The athletes were fitted with the belt and accelerometer and then asked to perform 10 or more consecutive bounces on each of the three trampolines. The athletes were instructed to bounce as high as they could control. A self-selected number of initial bounces were undertaken and the athlete announced verbally when he or she was bouncing maximally. Sampling was undertaken throughout all bounces similar to previous procedures (Briggs, 2014; Harden & Earnest, 2015). The ten bounces with the highest and most consistent sequence of accelerations were used as the bounce trials to characterize each trampoline's acceleration profile. Figure 3. Super Tramp trampoline. Data analysis: Following data capture and storage, PASCO Capstone software was used to extract relevant information from each bounce (Figure 4) (Shanahan, 2004). A bounce was defined as the period from trampoline bed contact to departure. The variables of interest for this study were: • time from start of a bounce to the peak acceleration, • time from peak acceleration to the end of the bounce, • total time of the bounce, • jump flight time, • jump height, • peak acceleration, • average acceleration, • average force and allometrically scaled force. Science of Gymnastics Journal 223 Science of Gymnastics Journal Sands W.A. et al.: COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES. Vol. 11 Issue 2: 223 - 237 ^Max.: 91.9 ¡[siT^I Run #1 M Mean: 54.0 K1 , Area: 8.0 nys* s EhI Bounce Star | j Bounce End Vs Time (s) Figure 1. Example of the data and variables that were extracted from the bounce accelerations. The multiple trials (i.e., bounces) were displayed using the Capstone software and a cursor was passed through the data to acquire the timing of the start of the bounce, time of the end of the bounce, peak acceleration time and peak acceleration value. Resultant accelerations were used for all analyses. The trials data were assessed for reliability via trends across trials (Henry, 1950, 1967). The means of the trend-free trials were calculated for each athlete collapsing the ten trials per trampoline-type to a single mean value which was later used for magnitude-based inference and hypothesis testing (Henry, 1950, 1967; Hopkins, Hawley, & Burke, 1999). The large number of performance trials (10 per athlete per trampoline-type) led to using Cronbach's alpha procedures to calculate an intraclass correlation coefficient (ICC) - alpha (Atkinson & Nevill, 1998). Additionally, one-way repeated measures ANOVAs were calculated across the ten trials along with coefficients of variation (CV) for each variable obtained from each trampoline-type (Table 1). Nine variables showed extremely high ICCs while also indicating some statistical differences across trials (Table 1). Closer inspection of these data showed no consistent pattern of variability such as increasing values indicative of learning or decreasing values indicative of fatigue. Therefore, because the ICCs were extremely high, CVs were low or modest, a reluctance to discard data (Henry, 1950), and no apparent pattern of variations across trials, all data were retained and means were calculated utilizing all ten trials for each athlete and each trampoline-type. Our initial assessment involved calculating reliability and trends across trials values and coefficients of variation with sex as a group factor. After reliability assessments, multiple 9 (variables) by 2 (sexes) by 3 (trampoline-types) repeated measures ANOVAs (RMANOVA) were calculated. The data showed that there were no main effect statistical differences attributable to sex (all p > 0.05). Following these uniform results, the data were collapsed across sex and further analyses involved multiple (9 variables) one-way RMANOVAs. All data were analyzed using IBM SPSS software (IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: Science of Gymnastics Journal 223 Science of Gymnastics Journal Sands W.A. et al.: COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES. Vol. 11 Issue 2: 223 - 237 IBM Corp). Effect size estimates were — small, 0.02 to 0.13 — medium, 0.13 to calculated as partial eta2 (q2) values: < 0.02 0.26 — large (Cohen, 1988). Table 1 Trials Reliability. Variable Trampoline Chronbach's RMAnova Coefficient of Standardized F(9,8i) P Variation Item Alpha Mean (SD) Time Start to Peak Tumbl Trak .96 1.15 .34 18.88 (1.18) Standard .97 1.69 .11 11.82 (11.09) Super Tramp .96 0.50 .87 9.76 (0.76) Time Peak to End Tumbl Trak .99 0.91 .53 10.16 (4.96) Standard .97 1.17 .32 10.93 (5.47) Super Tramp .96 1.06 .37 11.46 (6.26) Total Time Tumbl Trak .98 0.68 .73 11.33 (3.73) Standard .96 1.06 .40 10.13 (2.79) Super Tramp .94 0.72 .79 8.97 (2.87) Flight Time Tumbl Trak .99 0.50 .85 4.10 (2.52) Standard .98 4.47 <.001 3.42 (1.97) Super Tramp .99 0.95 .48 2.71 (1.02) Jump Height Tumbl Trak .99 0.50 .85 8.13 (4.92) Standard .99 4.95 <.001 6.82 (5.40) Super Tramp .99 0.88 .54 5.39 (2.02) Peak Acceleration Tumbl Trak .97 1.55 .14 8.14 (4.92) Standard .99 7.97 <.001 6.83 (3.91) Super Tramp .99 2.72 .008 5.40 (2.02) Average AccelerationTumbl Trak .97 3.12 .003 6.11 (2.07) Standard .98 1.13 .36 5.24 (2.52) Super Tramp .99 2.64 .010 3.80 (2.14) Average Force Tumbl Trak .99 3.40 .001 6.12 (2.07) Standard .98 1.23 .29 5.24 (2.52) Super Tramp .99 2.48 .015 4.01 (2.26) Average Force Tumbl Trak .97 3.12 .003 6.11 (2.07) Allometrically Scaled Standard .98 1.13 .36 5.23 (2.52) Super Tramp .99 2.64 .01 4.01 (2.26) Science of Gymnastics Journal 223 Science of Gymnastics Journal Sands W.A. et al.: COMPARISON OF BOUNCE CHARACTERISTICS ON THREE TYPES. Vol. 11 Issue 2: 223 - 237 RESULTS Table 2 One-way Repeated Measures ANOVA. Variable F df P Effect Size „2 I| partial Power Time Start to Peak 43.10 2,18 <.001 0.82 1.00 Time Peak to End 6.00 1.17,10.49 .030 0.40 .64 Total Time 31.35 1.27,11.40 <.001 0.78 1.00 Flight Time 248.72 1.38,12.41 <.001 0.97 1.00 Jump Height 159.65 1.27,11.43 <.001 0.95 1.00 Peak Acceleration 5.34 2,18 .015 0.37 .77 Average Acceleration 2.20 2,18 .140 0.20 .39 Average Force 2.19 2,18 .140 0.20 .39 Average Force Allometricallv Scaled 2.20 2,18 .140 0.20 .39 SUP Total Time STD Total Time TT Total Time