Anthropometric characteristics vary by game position and demonstrate correlation with motor performance in handball

Aim: The aims of this study were to describe and compare anthropometric variables and motor performance between playing positions, and to test the correlations between these variables in adult handball players. Methods: Twenty-three handball players (20,78 ± 3,83 years) participated in the study, being subdivided by position into backs (n = 9), wings (n = 9) and pivots (n = 5). Two assessment batteries were carried out, with an interval of 72 h. In the first battery the anthropometric (height, body mass, lean mass, fat mass, and fat mass percentage) and performance variables were analyzed in the squat jump, counter movement jump, and standing broad jump. In the second battery, the performance in the 10 m sprint test and agility were evaluated. Results: The pivots showed higher body mass, fat mass and fat mass percentage when compared to backs and wings, with pivots and backs being the tallest players in the team (p < 0.05). The performance of the pivots in the T-Test was lower than the backs. Correlations were found between high fat mass and low performance in the T-test, squat jump, counter movement jump and standing broad jump, and between high fat mass percentage and low performance in the T-test, squat jump, and counter movement jump. Conclusion: Anthropometric characteristics and motor performance vary depending on playing position. In addition, moderate correlations were found between high fat indices and low performance on agility and vertical and horizontal jumping tests. Key-words: Exercise, Physical functional performance, Velocity measurement. Artigo Original Revista Brasileira de Fisiologia do Exercício ISSN Online: 2675-1372 ISSN Impresso: 1677-8510 RBFEx 360 Oliveira LP et al. Antropometria e desempenho motor no handebol


Introduction
The anthropometric characterization of handball players has been an object of interest in several studies in the sports sciences and training areas [1][2][3][4][5][6][7]. The relevance of this knowledge is the possibility of; 1) collaborating with technical staff in the search for "talents" in this sport, due to the adequacy of athletes with respect to the desired to anthropometric profiles, and; 2) To assisting professionals, such as fitness coaches to better outline their training programs in order to achieve the intended body profile for a player in this sport modality.
It is documented in the scientific literature that elite level professional handball players have greater height (H) and body mass (BM) when compared to amateurs [6] and non-elite professional [7] players, and that these differences also exist between the best and worst teams ranked teams in renowned world championships [8]. Similarly, it has also been identified that elite professional players have higher absolute lean mass (LM), and lower amounts of absolute fat mass (FM) and body fat mass (BFM) [7].
The comparison of motor skills has also been a target of interest in studies in this field [6,7]. Differences have been shown in motor tests applied to these players, with special emphasis on the 10, 20, and 30 m [7,9] sprint time, absolute power of the lower limbs in the squat jump (SJ) and counter movement jump (CMJ) [7], and ball speed in the throw [6,9]. In this context, more recent studies also point out that the physical characteristics depend on the specific game position of the player, subdividing them into backs, wings, pivots, and goalkeepers [10][11][12][13].
Although anthropometric variables have already been the target of research that found differences between game positions [11][12][13][14], and that some motor tests were not sensitive to detect differences in physical performance [12], little is known about the real influences of anthropometrics variables on motor tests in handball players taking into account the game positions.
Thus, the main objectives of this study were 1) to describe and compare anthropometric characteristics and motor performance between game positions; 2) to test correlations between anthropometric variables and motor performance in adult handball players.
Additionally, the main hypothesis of the present investigation was the possibility of motor performance varying by position according to the higher or lower requirement of certain motor skills (e.g., force, velocity, and resistance), as well as that anthropometric variables could be sensitive to explain differences in motor performance.

Participants
Twenty-three adult handball players (20.78 ± 3.83 years; 86.98 ± 15.95 kg; 182.69 ± 6.75 cm; 16.19 ± 7.39%) were considered as the sample population, of amateur level, belonging to a team that disputes the first division of the Paulista championship. The inclusion and exclusion criteria were having at least one year of uninterrupted training in the modality, and not presenting musculoskeletal injuries or health problems. For the analyses between the game positions, the players were subdivided into backs (n = 9), wings (n = 9), and pivots (n = 5).
The effect size of the sample size used in the present study (n = 23) was 0.53, calculated using G*Power software (v. 3.1.9.4), assuming α = 0.05 and β = 0.89, based on a previous study with handball players [15] using as a parameter for the calculation the level of correlation between body mass and performance in the throwing test (r = 0.53).
The collections were carried out at the beginning of the preparatory stage (February/2014) for the Paulista championship, a period in which athletes trained with a frequency of three sessions per week, lasting approximately two hours per session. This study was approved by the Research Ethics Committee of the Clinical Hospital of Ribeirão Preto (Protocol 775.212), Faculty of Medicine of Ribeirão Preto, and was conducted in accordance with the principles established by the declaration of Helsinki.

Experimental design
After being informed about the procedures adopted in the research and signing the Free and Informed Consent Term, the individuals were submitted to two batteries of assessments, carried out on an official handball court, on two assessment days, with a 72 h interval between them. The test days were divided into (I) anthropometry (H, BM, LM, FM, and BFM), and motor tests (CMJ, SJ, and SBJ), and (II) motor tests (10 m sprint and T-test).

Anthropometry
The variables BM and H were assessed using a digital scale (DLK Sports, SB-623, Brazil) and digital laser measuring tape (Bosch, DLE-40, Germany), with precisions of 0.1 kg and 1.5 mm, respectively.
To estimate the BFM and the amount of LM and FM, first, the body density (BD) was estimated using the equation proposed by Jackson and Pollock [16], based on the sum of three skin folds, chest, abdomen, and thigh (X 2 ) and age in years (X 3 ) (Equation 1). A scientific adipometer (Sanny, AD1010, Brazil) with a precision of 0.1 mm was used for the measurements, following the procedures proposed by Harrison et al. [17]. DC = [1.1093800 -0.0008267 × (X 2 ) + 0.0000016 × (X 2 ) 2 -0.0002574 × (X 3

)] (Equation 1)
To estimate BFM based on body density, the equation proposed by Siri [18]

Velocity test
For the velocity test, the participants were positioned behind the start mark and, after a sound signal, they covered the distance of ten meters in a straight line in the shortest time possible. To record the time taken in the test, one evaluator was positioned at the starting point, and another at the end of the ten meter distance, the first being responsible for signaling the start of the test and the second for marking the time spent using a stopwatch (DLK Sports, WT-038, Brazil), accurate to 1/100 seconds.
Oliveira LP et al.

Anthropometry and motor performance in handball
Agility test For the agility T-test, adapted by Moreira, Souza and Oliveira [19], the participants started the test positioned behind the start mark and after a sound signal, they covered ten meters in a straight line, until the first cone (central cone). After touching the cone with their hand, they changed direction to the left, towards the next cone (positioned at five meters in relation to the central cone). The participants then returned in the opposite direction, moving to the other cone of the "T", covering another ten meters. Finally, the subjects returned to the central cone, covering five more meters, and then finished the test by covering another ten meters towards the finish mark, totaling 40 m in the test (Figure 1).
To determine the time taken to perform the adapted T-test, a stopwatch (DLK Sports, WT-038, Brazil) was used, with an accuracy of 1/100 seconds.  [19]).

Vertical jump test
The vertical jump test was conducted on Ergo Jump equipment (Cefise®, Brazil) associated with the Jump System Pro software (Cefise®, Brazil), version 1.0. The mat is composed of electronic circuits which enable the estimation of vertical jump height and power of the lower limbs based on the flight time and acceleration due to gravity [20].
During the test, participants performed two execution techniques: Squat Jump (SJ) and Counter Movement Jump (CMJ). For SJ, the participants positioned themselves on the mat in a standing position, with their feet parallel, and their hands on the hips, in order to neutralize the action of the upper limbs. The individuals were instructed to start with a 90º knee flexion and perform a vertical jump to the highest possible height. For the CMJ, the positioning of the feet followed the same procedure adopted for the SJ. For the jump, a flexion movement was performed followed by knee extension, starting from an upright position [21].

Standing broad jump test
The standing broad jump test was performed on the court where the athletes habitually train. The participants started the test behind the exit mark, with their feet slightly apart, and were instructed to perform a semi-flexion of the knees to-gether with oscillatory movements of the arms and to jump to the greatest possible horizontal distance [22]. The jump distance was measured using a measuring tape (Stanley, 34-263, United States) fixed to the floor of the court. The longest distance was measured between the starting line and the heel closest to the starting mark.
For all tests in the study, three attempts were performed, and the best attempt was considered for statistical analysis. The recovery time between tests was five minutes and one minute between attempts in the same test [21].

Statistical analysis
Initially all the data passed the Shapiro Wilk normality test, which allowed parametric statistical analysis. After confirmation of the normality of distribution, one-way ANOVA analysis was adopted, followed by the Tukey post hoc to compare anthropometric variables and motor performance between positions (i.e., backs, wings, and pivots). Possible correlations between the variables were tested using Pearson

Results
Table I presents the mean values of the anthropometric variables and motor performance, as well as the comparisons between the game positions. The pivots demonstrated greater BM, FM and BFM compared to the backs and wings, and the pivots together with the backs were the tallest in the team (p < 0.05). Regarding motor performance variables, the pivots presented a worse time in the T-test compared to the backs (p < 0.05). In contrast, no significant differences were found in the 10 m, SBJ, SJ, CMJ, and absolute mean power (MP) and relative power (RP) of the lower limbs for both SJ and CMJ techniques.
Considering the total sample (n = 23), Table II presents the correlation values between the anthropometric variables and motor performance. A positive correlation was observed between BM and T-test time and, an inverse correlation with the SBJ, SJ, and CMJ. The FM presented a positive correlation with the T-test, and inverse correlation with the SBJ, RP in the CMJ, and vertical heigth in the SJ and CMJ. For the BFM the same behavior was observed, a positive correlation with the T-test, and inverse correlation with the RP in the CMJ, and vertical heigth in the SJ and CMJ, highlighting an association between lower values for these variables (i.e., BM, FM, and BFM) and better performance.

Discussion
The main findings of this study were the diferences in antropometric variables between the positions of backs, wings, and pivots and also in agility test performance between backs and pivots. In addition, significant correlations were found between anthropometric variables and motor performance.
Players occupy different positions on the court, in which they are required to perform functions directly related to the model and game system adopted by the coach. The backs, for example, are positioned in more distant places from the opponent's goal, allowing displacements in different directions of the court and greater distance in relation to their direct and/or indirect markers. The backs can also use the greatest number of technical-tactical elements (such as crossings, feints, and changes of direction) in an attempt to obtain advantageous situations to attack and allow infiltrations or finishings of medium or long distance by other backs [24]. The wings are players who act close to the right and left side lines of the court (usually close to the end lines or nine meters lines on the court). These players perform the functions of initiating the movement of the ball and finishing after that movement (in a positional attack), developing different collective technical-tactical elements (such as crossings, changing of specific positions, and curtains), participating in changing of offensive systems (when they occupy the pivot post -or second pivot), and also initiating counterattacks that provoke defensive imbalances [24]. The pivot is the striker positioned closest to the opponent's goal and among the defenders, whose body position is usually lateral to or their back to the opposing team's goal [25]. Despite the apparently fixed positioning, the pivot performs actions such as blocking, which make it difficult for defenders to move and makes it possible for infiltration by back players, and to clear them, so that they can receive the ball and perform the spin for the throw, which requires high strength levels of this player [25].
When analyzing the demands of the game and the functions mentioned above, it is possible to suggest that in handball each position occupied on the court requires certain anthropometric characteristics of the players. For example, the physical attributes of the pivots, such as the greater H and BM found in this study, may be favorable in offensive situations, such as in blocking actions, to facilitate infiltration and finishing for backs and wings. On the other hand, the greater agility (i.e, better T-test time) of the backs and the lower BM and BFM of the backs and wings, when compared with pivots, could facilitate feints, changes of direction, and quick offensive actions, which can be developed together with other players such as the pivots.
Regarding the performance in motor tests, the main hypotheses were the possibility of performance varying according to the higher or lower requirement for a certain motor skill in the position, and that the anthropometric variables could be sensitive to explain differences in performance. In this sense, significant differences were found in the agility test, with the pivots presenting a worse time in the T-test compared to the backs. In addition, negative correlations were found between BFM, FM, and performance in the T-test, SBJ, SJ, and CMJ.
Unlike anthropometric aspects, few studies in the literature have investigated motor performance taking into account game positions [12,13,26]. Chaouachi et al. [12] analyzed 21 professional handball players of elite level and pointed out that there were no significant differences in motor performance (i.e., jumping, sprint, upper and lower limbs strength, throw speed and aerobic power) between goalkeepers, backs, pivots and wings. On the other hand, Massuça et al. [26] recently demonstrated, in a sample composed of 161 handball players, including professionals and non-professionals subdivided by position, significant differences in the 30 m sprint test, MP of the lower limbs, and handgrip strength. Therefore, according to the authors, motor performance may also vary depending on the game position. Thus, it is possible that in the present study the reduced sample number may have provided a low statistical power to observe significant differences between the game positions for the other variables (10 m, SBJ, CMJ, SJ, MP, and RP).
Regarding the correlation data, Dellagrana et al. [27] and Mota and Virtuoso Junior [28] found results similar to those of the current study, but in young handball players and university students, respectively. Dellagrana et al. [26] identified an inverse relationship between the BFM and SBJ (r = -0.42), and a positive relationship between the BFM and time in the Shuttle run test (r = 0.61). Mota and Virtuoso Junior [27] found an inverse relationship between BFM and maximum oxygen consumption (r = -0.55), estimated from the Balke test performed on a bicycle. Thus, the correlation of the results of this study with the data available in the literature shows that anthropometric variables have an influence on motor performance, and that the magnitude seems to be dependent on the test employed.
Muscle tissue produces strength actively during the process of muscle contraction, through the cross bridges (formed by the myofibrils actin and myosin), contributing to the performance of motor gestures. Adipose tissue, on the other hand, is not able to produce strength actively, therefore, it is possible that the excess body fat from the pivots may have prejudice the physical performance during the T-test, mainly due to the significant correlations found for FM and BFM, and the fact that the pivots demonstrated higher values than the other positions in these two variables. In addition, the fact that there are no significant differences in RP between positions, suggests that even players with higher levels of FM and BFM, such as pivots, maintained a good rate of muscle power/body weight, allowing similar performances in tests with high demands on motor strength and velocity (e.g., 10 m sprint, vertical jump, and standing broad jump).
In summary, based on the results found and on the aforementioned studies, although only specific differences were found in the performance for the agility test, there is the possibility that performance may be altered depending on the game position. In addition, the influence of anthropometric variables on motor performance was identified, with the magnitude appearing to be dependent on the motor test used.
As a limitation of the study, our experimental design involved only analyses using motor tests, which has few implications for the actions that occur during the context of the game, especially with respect to the movements of the SBJ, CMJ, and SJ tests, which can be considered as general and not specific in relation to the sport modality. In addition, the measurement instruments used (e.g., jumping mat, stopwatch) despite having low methodological complexity and high practical application for athletes and coaches, present large systematic measurement errors. Therefore, it is suggested that more robust analyses of physical and tactical performance by game position should be performed, using better measurement instruments and that include the analysis of displacement patterns during a handball game.

Conclusion
The findings of this study suggest that for adult handball players of amateur level, anthropometric characteristics vary depending on the game position. The results showed that the pivots have higher BM, FM, and BFM when compared to backs and wings, and that pivots and backs are the tallest players on the team. Motor per-formance also varied between positions, and it was found that backs achieved better T-test performance when compared to pivots, while no significant differences were observed in the other motor tests (10 m, SBJ, CMJ, SJ, MP, and RP). Finally, moderate correlations were found between FM and BFM with performance in the T-test, SJ, CMJ, and SBJ, which suggests that these factors influence, even if in a small measure, motor performance.

Potential conflict of interest
No conflicts of interest with potential potential for this article have been reported.

Financing source
There were no external sources of funding for this study.