EVALUATING PERFORMANCE PROGRESSION IN BEGINNER WHEELCHAIR RUGBY

The purpose of this study was to describe the change in selected fitness and skill components over a nine-month period of a new intensive Wheelchair Rugby (WR) practice program. Sixteen WR players of the newly established national league participated in this intervention, which included team and individual practice sessions. Five items of the Beck Battery of Quad Rugby Skills and a 10-min wheelchair push test were performed at the beginning (t 1), after three months (t 2), and after an additional six months period of practice (t 3). Based on repeated measures t-tests for mean data of at least 50% of the participants, significant improvements were found in the 10-min and in the sprint tests between t1 and t 2, and in the picking and manoeuvrability tests between t 2 and t 3. It can be concluded that this training program was effective in improving players' endurance and manoeuvrability. The application of the test battery for training design and progression control appears very useful.


INTRODUCTION
Wheelchair Rugby (WR), previously called Murder ball and Quad Rugby, is one of the most recent developments in Paralympic sports.It is a team sport for male and female persons with a disability in both their lower and upper limbs, primarily due to spinal cord injury (SCI).In the recent years more and more athletes with other disabilities affecting all four limbs have joined WR teams.WR was invented in 1977 in Winnipeg, Canada by athletes with cervical SCI, and combines elements of basketball, handball, and ice hockey (International Wheelchair Rugby Federation, 2010a).The object of the game is to carry the ball across the opposing teams' goal line, while demonstrating firm control of the ball.WR is played by two teams composed of four players each on a basketball court.It is an equal access, co-gender sport, and therefore females are playing together with males at the same teams.Internationally WR is governed by the International Wheelchair Rugby Federation, (IWRF, 2010b), who determined that the total number of points of all four athletes actually playing on court at any time cannot exceed 8.0 classification points (IWRF, 2010b).Since 1991, a functional classification system has been used to qualify athletes for the game, with seven classes of 0.5 through 3.5 points, accommodating the growing number of athletes both with spinal cord injury and without (such as polio, cerebral palsy, muscular dystrophy, multiple sclerosis and amputations) (IWRF, 2008).Athletes in the higher classes, or "minimal" disability, who are eligible for the sport of WR are more functional and usually play offensive roles, while those in the lower classes typically play defensive roles that require different wheelchairs.Since the Sydney Paralympics in 2000, WR has acquired a full medal status at summer Paralympic Games (2010a).According to the World Ranking List, effective November 9, 2009, 24 teams are ranked and 10 are recognized as representing developing countries (IWRF, 2009), one of which is Israel, where this game has been under development since mid 2009.

Conditioning Requirements in Wheelchair Rugby
Wheelchair Rugby is an intermittent type of sport game, with many skilled activities that share similar tasks as in wheelchair basketball (e.g., wheelchair propulsion, manoeuvring, dribbling, passing, and blocking).However, due to the upper limb involvement the patterns of these skills are different than in basketball.Also, WR is more aggressive than wheelchair basketball, and players are allowed to collide in each other.Players are allowed ten seconds with the ball while pushing the wheelchair, before having to pass or dribble.Therefore this pattern of up to 10 seconds of pushing characterizes most of the repetitive powerful actions required on court.
The pushes are often asymmetrical due to the need to escape from offensive blocking, and require a high degree of explosive strength and anaerobic fitness.
In addition, the accumulated effort during the period played usually challenges aerobic fitness (Morgulec, Kosmol, Molik, Huber-Wozniak, & Butkowska, 2006).A medium significant correlation (r= .72,p<.05) found between mean power output and VO 2peak of WR players in an armcranking sprint test (Goosey Tolfrey, Castle, & Webborn, 2006) demonstrates the relationship of aerobic and anaerobic performance in WR.Yilla and Sherrill (1998) constructed and validated a specific WR skill test labelled the "Beck Battery of Quad Rugby Skill Tests".This battery included five tests dedicated to the measurement of performance variables (labelled as skills) typical for WR, including sprinting, manoeuvrability with the ball, picking, passing for distance, and passing for accuracy.After analyzing the data of 65 participants using an oblique rotation, a twofactor solution was revealed, with one factor comprised of manoeuvrability, sprinting, and passing for distance, and the other factor of passing for distance, passing for accuracy, and picking.Nevertheless, it should be noticed that both factors included components of wheelchair as well as ball handling.Both factors explained 66% of the total variance.Factor loading for manoeuvrability with the ball (r=.99) was highest of the five items, and the authors recommended that this should be the test of choice for assessing players' performance.This test, together with sprinting and picking, were included in a study examining the relationship of these tests with an anaerobic arm ergometry test (Morgulec & Kosmol, 2007;Morgulec-Adamovicz, Kosmol, & Molik, 2010).Good correlations ranging between r= .60 to .84 were established, suggesting that these field tests lasting 6 -100 sec were strongly related to anaerobic performance.
In addition, the effects of one year of training on aerobic performance in arm cranking have been reported, describing a significant increase in all measured variables, including VE peak ; VO 2peak ; and test time (Morgulec et al., 2006).However, until now the sensitivity of fitness performance variables measured in the field to training intervention has not been established.The purpose of this study was to describe the change measured in selected fitness and skill components over a nine-month period of an intensive WR training intervention.A secondary purpose was to establish the factor structure, internal consistency, and validity of the modified test battery.

Method Participants
Participants in this study were 16 WR players who comprised the teams playing in the first season of the newly-established Israeli league.Participants' age ranged from 14 years to 38 years of age (Mean = 26.3,SD = 6.8 years).Of the 16 players, 12 had a spinal cord injury (SCI), two had congenital L-Chad non-progressive muscular dystrophy, one had Charcot Marie Tooth hereditary motor and sensory neuropathy, and one had polio.For the participants with SCI, the number of years since injury ranged from one to 18 years (Mean = 6.8,SD = 5.9 years).Fifteen players were male and one was a female.In the absence of an official international classification, the players' classification was estimated before performing the tests based on the assessment of two experts (a coach and an international Berzen, Hutzler Evaulating wheelchair rugby 55 EUJAPA, Vol. 5, No. 1 wheelchair sport scholar) who had read the classification manual (6).The sample included a variety of estimated classifications between 0.5 and 3 points.No 3.5 point assessments were reported in this sample..  Sherrill (1998).Based on preliminary experimentations our players, who were beginners, had difficulties to cope with the structure of the pass for accuracy test.Therefore, it was modified in the shape of the target (see Figure 1) compared to the original circular shape, and the number of scores (up to 9 rather than 10 points).For the similar purpose, the manoeuvrability test was also modified from the original.The modified manoeuvrability test was carried out as follows: Players performed slalom in and out of cone gates (Five 3-m and four 2-m gates) positioned in a row, from one end of the course to the other, while maintaining a legal dribble of 10 seconds.If the ball was dropped the player got a zero for that trial, unless he or she was able to retrieve the ball and continue from the point it was dropped (no assistance was allowed).The time was recorded when the player crossed the end line.Two trials were given to each player.In addition, a 10min test for maximal distance was performed to provide a measure of endurance.The five Beck tests were performed in one training session and the 10-min test in another session during the same week.were able to attend all three testing sessions, due to illnesses and availability barriers.
Between 10 and 14 participants were present at all test times, with the exception of manoeuvrability at t3 (only 6 athletes).

Training Intervention
The activity program changed gradually during the progression of the program from predominantly aerobic, strength build-up, and wheelchair propulsion oriented during the first three months of the program (between t1 and t2), to more focused on power in each of the specific WR skills, techniques, and tactics, relative to the competition schedule, which took place during the second half of the period between t2 and t3.Throughout the training program participants attended two practice sessions per week lasting about two Normality of data was analyzed using the histogram method.
In order to assess the test structure, a principal component factor analysis with Varimax rotation was performed on all data points of each test.In addition, a correlation was performed between all data points across the test times for repeatability.

Results
Table 3 presents comparisons of means and standard deviations (SD), correlations of the repeated measures of participants attending t1 and t2 as well as t2 and t3, and their effect sizes across tests and times.In addition, change scores between t 1 and t 2 and between t 2 and t 3 were calculated, and represented as a percent change of the test performed at the previous time (Table 4).For example, the percent change for pair Manouev1 and Manouev2 presents the percent performance aggregated to the performance in Manouev1.Using the Bonferroni correction for repeated pair comparisons, a significant improvement (p<.025) was found for mean data of at least 50% (eight cases) of participants in the 10min test between t1 and t2, and in the picking test between t2 and t3.Also, the manoeuvrability test demonstrated a substantial change (23.2%) between t2 and t3 (see Table 4).However, it should be noted that only four cases of repeated measures have participated in this test sample, which was apparently not enough for reaching significance or generalizing for the whole sample.Medium effect sizes around the 0.5 level (2) were observed only in the picking and manoeuvrability tests between t2 and t3.The principal component factor analysis performed on all data points (n=48) revealed a two factor solution (Eigenvalues > 1), explaining 82.5% of the variance.The first component explaining 58% was composed of the three power tests manoeuvrability, sprint and picking (rotated item loadings of .95,.93,and .91,respectively).The second component with 24.5% of the variance was composed of the other variables 10-min push, pass for accuracy, and pass for distance (.86, .70, and .49 loadings, respectively).Figure 2 presents the visual distribution of items within the two dimensional rotated space, and demonstrates the strong relationship of the three power items.

Training outcomes
A Dutch group (Dallmeijer, Hopman, van As, & Van der Woude, 1996) evaluated the effect of WR training over a period of six months, on fitness components and physical strain in performing daily activities, such as wheelchair transfers and opening a door.They reported a significant increase in isometric strength and in maximal anaerobic and aerobic performance, and a reduction in physical strain during activities of daily living, due to the WR training.These objective outcomes were reinforced with questionnaire data from New Zealand (Adnan, McKenzie, & Miyahara, 2001), suggesting an impact on daily activities such as entering and leaving a car, and transferring to and from the wheelchair.In addition, data emerging from a thematic analysis of in-depth interviews with WR players (Goodwin et al., 2009) indicated increased independence experienced through participation in the sport."I get way more independence by playing rugby then I ever did from a doctor…that's an absolute truth."(Goodwin et al., 2009, p. 110).Our findings support these previous findings that were conducted with both beginner and expert players.Our outcomes can be partially explained by the content of the training program throughout the period evaluated.Since the first period of three months concentrated more on aerobic and basic strength build-up, and the second period increasingly focused on power and WR skill and tactics training (see Table 2), it sense that during the first period the 10 min test results and in the second period the picking test results improved significantly.In addition, the manoeuvrability test results improved with a medium size effect during the second period.Our findings support previous studies following 6-month (Dallmeijer et al., 1996) and12-month (Morgulec &Kosmol, 2007) WR training programs, revealing significant improvement in endurance performance variables of WR players.Our findings are the first to document an improvement in outcomes of the picking and the manoeuvrability tests that can be associated with power and agility.Previous results indicate that these tests, together with the sprint test demonstrated medium to strong correlations (range r=.60 to .84) with anaerobic maximal and mean power in an armcranking 30-sec all-out test (Morgulec-Adamovicz, Kosmol and Molik, 2010).Based on armcranking measurements of anaerobic capacity (Morgulec, Kosmol, Vanlandewijck, & Hubner-Wozniak, 2005), as well as outcomes of self-efficacy evaluations for activities of daily living in WR players compared to non participants with high level spinal cord injury (Adnan et al., 2001), the WR players exhibited significantly higher anaerobic capacity, and reported significantly better self-efficacy in activities requiring power and agility, such as transferring to and from wheelchair to bed, leaving / entering a car, and transferring from seat to wheelchair.These performance variables appear to have an important impact on the well-being in the daily life of WR players.Based on the improvement gained in test performance of our participants after introducing the training program, it may be suggested that this activity could enhance daily life functions.Future studies are encouraged in order to measure the relationships between selected WR and daily life performance measures.

Test structure
The pooled outcomes of players' performance in the six test items selected provided a robust factorial solution, explaining 82% of the variance.The six test items were divided into two components, with one of them clearly characterizing power and explosive skills required in WR.Yilla and Sherrill (1998) revealed a similar factor structure for the much larger sample of 65 adult male players, and called this factor manoeuvrability with the ball.They recommended that manoeuvrability should be the test of choice, if players can only be tested on one item.In our sample this test also revealed the highest loading within the test items (.955).Nevertheless, such a practice would omit the ball handling component and therefore should not be advised.The repeatability of the test items across the different testing times was high with correlations for five of the tests ranging between .84 and .99,and moderate for the pass for accuracy (r ranging between .51 and .73).

Limitations
This was a test of one group of participants, who were at beginner stage of performance.It may not apply to players who are more trained and experienced.In addition, we had a number of missing values that reduced the generalizability of outcomes within our own sample.

Conclusion
This study revealed significant improvement in WR performance, particularly in the endurance, power, and manoeuvrability-related activities, following a 9-month period of training intervention.The test items used in this study present a potential for future use, but require continuous evaluation of their structure and reliability within studies that include larger samples.

Figure 1 .
Figure 1.Description of Throwing for Accuracy test setting

*Figure 2 .
Figure 2. Component plot in rotated space depicting factor loadings of tests

*
Israel Sport Center for the Disabled, Israel ** Zinam College for Physical Education and Sport Science

Table 1
presents the individual data -Personal data of participants in the study

Table 2 .
Example for a training plan

Table 4
Paired differences and significance of t-tests performed across tests and time