Testing methods and physical qualities of male age grade rugby union players: A systematic review

Background Rugby union match demands are complex, requiring the development of multiple physical qualities concurrently. Quantifying the physical qualities of age grade rugby union players is vital for practitioners to support athlete preparation and long-term development. Aim This systematic review aimed to identify the methods used to quantify the physical qualities of male age grade (≤ Under-20) rugby union players, present the normative values for physical qualities, and compare physical qualities between age grades and positions. Methods Electronic databases were systematically reviewed from the earliest record to November 2019 using key words relating to sex, age, sport and physical testing. Results Forty-two studies evaluated the physical qualities of age grade rugby union players. Seventy-five tests were used to quantify body composition, muscular strength, muscular power, linear speed, change of direction ability, aerobic capacity and anaerobic endurance. Thirty-one studies met the eligibility criteria to present the physical qualities. Physical qualities differentiate between age groups below Under-16, while differences in older age groups (Under-16 to Under-20) are not clear. Positional differences are present with forwards possessing greater height, body mass, body fat percentage and strength while backs are faster and have greater aerobic capacities. Conclusions A wide variety of tests are used to assess physical qualities limiting between study comparisons. Although differences in older age grades are unclear, older age groups (Under-19-20) generally performed better in physical tests. Positional differences are associated with match demands where forwards are exposed to less running but a greater number of collisions. Practitioners can use the results from this review to evaluate the physical qualities of age grade rugby union players to enhance training prescription, goal setting and player development. Future research should consider the use of national standardised testing batteries due to the inconsistency in testing methods and small samples limiting the reporting of positional differences.


Introduction
Rugby union (RU) is a sport played in 121 countries, with over 8.5 million participants [1]. Professional competition is mainly based within Tier 1 nations from the southern hemispheres Rugby Championship (Argentina, Australia, New Zealand and South Africa) and the northern hemispheres Six Nations (England, France, Ireland, Italy, Scotland and Wales). During a senior level match, two teams of 15 players (with a maximum of 8 replacements totalling a 23 man squad) compete over two 40-minute halves with half time not exceeding 15 minutes [2]. Playing positions can be categorised into two positional groups, forwards (prop, hooker, lock, flanker and no. 8) and backs (scrum half, fly-half, centre, wing and full back) [3]. Specialist roles during match play determine the demands for each position with the forwards suggested to be ball winners while the backs are ball carriers [4].
To support the development of young RU players towards the senior game, many national governing bodies have established age grade development pathways culminating with the Under 20 (U20) World Championship. Developing technical, tactical and physical qualities throughout the pathway is key to enhancing RU performance [5][6][7][8]. RU match play is complex, with collisions, high speed running, and technical elements being interspersed with periods of recovery [9,10]. The match demands of RU players do not only vary depending on playing position but also age and playing level [11,12]. Over the course of an U20 International game forwards and backs are required to run~5370 and~6230 m, with~284 and~657 m covered at high speed (> 5.0 m�s -1 ), respectively [13]. To date there is no information regarding the velocities achieved during U20 International games however both age grade and senior players achieve velocities greater than 90% of their maximum sprint speed during match-play [14,15]. A similar number of collisions has also been observed across playing levels for both forwards (~40) and backs (~13), although the magnitude of collisions may differ due to differences in body mass [16,17]. Due to the complex and dynamic nature of RU multiple physical qualities need to be developed to optimise RU performance.
Although the physical qualities of senior RU players have previously been reviewed [10], only the height and body mass of age grade RU players are summarised in the current literature [18]. Throughout the development pathway it is important for practitioners to appropriately develop the physical qualities of players to promote optimum performance and long-To address the first research aim, studies were eligible for inclusion if they explored the physical qualities (anthropometrics, strength, power, speed, change of direction, aerobic capacity and anaerobic endurance) of male, age grade (� U20) RU players. The article was included if it identified at least one anthropometric or physical quality and this was the primary aim of the paper. Identification of these qualities as a result of other research aims (e.g. match demands, fatigue and nutrition) resulted in exclusion. Age was identified by either the reported age or the age grade stated in the article. Only articles written in English that appeared in a peer reviewed journal were included. If further information was required regarding the study, such as the age grade of players, the corresponding author was contacted and if there was no response the article was excluded.
For the second aim, common tests were identified for study inclusion and only studies that clearly identified the physical quality of a single age grade (e.g. U13 or U16) and not spanning a range (e.g. U17-U20) were included. If the study did not report team or positional means and reported multiple groups as part of an intervention it was excluded.

Assessment of methodological quality
The Downs & Black [27] assessment scale was modified to review the methodological quality of included articles. Due to the cross-sectional nature of the data extracted, similarly to previous research [28], only 12 questions (1-4, 6, 7, 10-12, 16, 18, 20) were used which were logically relevant. Due to no intervention taking place question 4 was interpreted as "are the tests in the study clearly described". The reporting of effect sizes was deemed acceptable for question 10. Failure to meet the criteria resulted in a score of "0" with sufficient information resulting in a score of "1". No studies were removed as a result of poor methodological quality.

Data extraction
The initial extraction withdrew publication data (authors, year of publication and sample information), all tests performed to quantify anthropometrics or physical qualities (e.g., onerepetition maximum (RM) back squat, countermovement jump, 10m sprint), and output measures reported (e.g., body fat percentage, jump height, sprint time). If the methods of physical tests could not be identified from the article or refences provided the tests were not included in the extraction.
Following the extraction of the tests used, data relating to the age grade, level of competition and playing position of the participants were noted along with results for anthropometrical (height, weight and body composition), strength (bench press and squat variations), power (vertical jump height, countermovement jump height and peak power), speed (10m, 20m, 30m and 40m), change of direction (Illinois agility test, T-test, pro agility and 505) and aerobic capacity tests (multistage fitness test (MSFT), Yo-Yo endurance test level 1 (YYE1), Yo-Yo intermittent recovery test level 1 (YYIR1) and 30-15 intermittent fitness test ). When needed WebPlotDigitiser V4.1 was used to extract means and measures of variance (standard deviations or confidence intervals) from figures [29]. When cross-sectional data for a single age group was presented for multiple years, the most recent time point was used. In the case of intervention studies, the baseline score was extracted to remove bias of the intervention.

Statistical analysis
The data are reported as a mean with a measure of variance as provided in the article. No further statistical analysis was carried out on the data.

Identification and selection of studies
The search identified 4,814 articles with a further 10 identified through hand searching reference lists. Following the removal of duplicates, 2,762 were screened and 149 studies were reviewed in detail (Fig 1). The initial extraction process found 42 studies assessing the physical qualities of age grade RU players with 31 studies used to present the physical qualities by age grade and position. Table 2 shows the characteristics of the studies that were included within the systematic review. Sample size ranged from 15-4007 with a median of 83 participants and the number of teams used to recruit the sample ranged from 1-188 with a median of 1. The age range of

PLOS ONE
The physical qualities of age grade rugby union players

Methodological quality
The score from the assessment of methodological quality can be observed in Table 3, ranging from 5-10 for the 12 items assessed.
Although grip strength testing [35,36,45,62,63] and flexed arm hang [34,36,44,50] were regularly observed throughout the literature, isometric testing methods were the least common with only wall sits [37,38] and isometric mid-thigh pull (IMTP) [40] added to this group. Flexed arm hang and wall sits were reported for a maximum time, while the grip test and IMTP measured peak force. Peak rate of force development was also reported for the IMTP [40].

PLOS ONE
The physical qualities of age grade rugby union players squat jump [47,64], 10 kg squat jump [53] and 60% 1RM squat jump [58]. Further tests of lower body power used in a single study included jumps in a horizontal vector (horizontal jump and triple hop) and 6s Watt bike peak power output [49,67,69]. Only two studies carried out tests of upper body power reporting peak power for bench throw (60% 1RM) [58] and distance for a 3 kg medicine ball throw [48]. Linear speed. Sprinting speed was the most reported physical quality shown in Table 6. Thirty-two of the forty-two (76%) studies reported speed related qualities. Time was the most common variable recorded over a range of distances from 5m to 60m (Table 4). Of the ten Table 6. Speed and change of direction tests and outcome measures reported by studies.
Aerobic capacity. Table 3 shows the sixteen (38%) studies that investigated the aerobic capacity of age grade RU players. The MSFT was reported as both number of stages completed [32,57] and estimated VO 2 max [30,35,36,47]. Additionally, other aerobic tests that were continuous in nature included the YYE1 [64] and 1500m run [41]. The YYIR1 [37,38,40,59,66,68,73] and 30-15IFT [39,40,73] provided an intermittent assessment of aerobic capacity reported as total distance covered and final running velocity, respectively. Anaerobic endurance. Only five studies (12%) tested the anaerobic endurance capabilities of age grade RU players (Table 3). Two tests reported the fastest time to cover 400 m [41] and 500 m [46], while two tests reported the distance covered during the 150 m [67] and 250 m [35] shuttle tests. Additionally, the Wingate anaerobic test was also used to assess anaerobic power through peak power, average power, total work and fatigue rate [48].
Body composition. Body composition was presented as a body fat percentage in sixteen studies ( Table 8). The highest body fat was recorded at U20 (22.1 ± 6.8%) [45]. All other age grades are reported to have a mean body fat percentage <16%, with the exception U12's [44] and one study at U16 [47]. The lowest reported body fat was observed in U16 provincial players (13.66 ± 4.77%) [42]. Apart from loose forwards at U13, all backs had a lower body fat percentage compare to forwards [32,34,49,52].
Anaerobic endurance. Due to there being no common tests within the literature, anaerobic endurance data was not considered in this section of the review.

PLOS ONE
The physical qualities of age grade rugby union players

Discussion
This is the first systematic review to outline the testing methods of physical qualities used in age grade RU research and present objective data of the physical qualities and compare between age grade and position. Following the screening process, 42 studies were found to measure physical qualities of age grade (� U20) RU players. From these studies, thirty-one were used to quantify the physical qualities of age grade RU players (U11 -U20) within the review, including; anthropometrics, body composition, strength, power, speed, change of direction and aerobic capacity. Although a large proportion of the data was not presented due to the variety in testing methods, more articles were identified to differentiate between age grades and position compared to similar reviews in RL and AFL [24,25]. The majority of research was cross-sectional, included players competing at U16 to U20 and was collected in Australia, Ireland, South Africa and the United Kingdom. It was identified in the review that a large number of tests are used within the literature to quantify physical qualities. When comparing between age grades and positions, differences are apparent and practitioners may find this information useful when evaluating the development of age grade RU players.

Methods of assessing physical qualities
The review of testing methods in age grade RU identified a total of 70 tests used to measure 7 physical qualities, demonstrating the wide variety of tests used within the research. In the case of power and speed, frequently used tests are easily identified (i.e. VJ, CMJ, 10, 20 30 and 40m). However, this was not the case for other physical qualities (i.e., body fat percentage, strength, change of direction, aerobic and anaerobic endurance). These findings are in accordance with previous literature where it is suggested the range of methods is due to the number of physical qualities important for the sport [21] and the origin of the country of testing [22]. Inconsistencies in the tests used could further be a result of resource constraints such as the availability of technology (e.g. VJ; vert-tec vs. CMJ; force plate or jump mat), time (e.g. 1RM strength testing vs isometric tests) or safety (e.g. 1RM strength testing vs isometric tests in untrained players). Furthermore, a change in the selection of tests over time may be influenced by a greater understanding of RU match demands where tests are selected to be more specific (i.e. sprint distance [14]). Finally, the development of scientific understanding and increased rigour of testing can influence test selection (e.g. validity of press up testing for strength [75]). These factors could be considered to represent a research-practitioner divide, where researchers favour scientific rigour and practitioners speed and cost of assessment. In the future it is therefore important for both practitioners and researchers to work together to design testing batteries to provide useful research and optimise evidence based practice [76]. As previously suggested in RL, there is potential demand for this connection to be made by Governing Bodies to introduce national/league wide testing batteries.
Even if a variety of tests were performed, multiple methods share common output variables. This can occur as a result of different equipment and techniques used for the same test (e.g. jump tests and height). Furthermore, calculations utilising the recorded value can be used to provide a new output variable (e.g. body fat percentage from sum of skinfolds or predicted 1RM from a multiple repetition maximum test). Although this increases the availability of data to be compared and used within practice, caution must be taken when comparing between methods. For example, methods for collecting body fat percentage have previously shown differing degrees of validity in professional RL players [77]. Furthermore, McMahon et al. [78] and Till et al. [79] have suggested it is not possible to confidently compare values between methods, providing specific examples for the CMJ and aerobic testing. In practice comparisons should first be made between studies that have used similar methods before selecting alternative sources.
A common theme throughout the literature is the use of absolute measures of performance with very few articles incorporating relative variables. With the large range of body shapes and sizes observed in RU it is important to consider the role of body size on athletic performance. This is especially important during age grade sport where increases in height and weight are observed during maturation [80]. To account for the effect of body mass, allometric or ratio scaling can be used to normalise strength and power where a larger muscle mass is beneficial [81,82]. In contrast, greater body mass is detrimental for speed performance but contributes to a higher sprint momentum which is shown to relate to collision success in International RU [6]. Although under reported in age grade RU relative measures of 30-15IFT [73] and speed performance [71] have shown to differentiate between age grade and senior RU players where absolute measures did not. Future research should therefore not solely rely on absolute measures of physical qualities but also report values relative to body mass.

Physical qualities
Anthropometric qualities. Anthropometric qualities (height and body mass) are important for RU performance due to the physical nature of the sport [83]. Both height and body mass are shown to increase with age, which is related to the process of growth and maturation [80]. Similar to longitudinal observations in RL, greater differences in anthropometry were found at younger age grades (i.e. U11-16) compared to older age groups as a result of growth and maturation [84,85]. There is a lack of research assessing the relationship between growth, maturation and the development of anthropometric qualities in age grade RU and further research is required to support this. Other factors that may influence a plateau in body mass within older age grades include the chaotic training demands placed on adolescent athletes in combination with low energy intakes compared to energy expenditure [86][87][88]. Although height and body mass are important for short-term performance it should be noted that Fontana et al. [47] found career success in RU was not dependent on height and body mass at U16. Furthermore, height does not differentiate between Elite U20 and Elite International RU players but there are differences in body mass (93.2 ± 12.3 vs. 102.8 ± 11.9 kg) [71]. It is therefore important to monitor the development of height and body mass throughout age grade RU as players go through maturation and develop towards the professional level.
When comparing between positions, forwards are generally both taller and heavier than backs, which are consistent with findings in senior players [10]. The size and mass of the forwards is better suited to the collision and set piece demands (i.e. scrum and lineout) they endure during a match [17]. Locks, also considered tight forwards, are the tallest at all age groups due to their role in the lineout where the goal is to maximise the peak height of the catch [89]. Furthermore, scrum halves are the lightest and smallest likely relating to the higher locomotive and reduced collision demands during match play in comparison to other positions [3]. Anthropometrics should be considered during the talent identification process and positional selection in combination with growth and maturation status.
Body composition. Body composition is an important consideration for performance as excessive body fat can negatively affect physical performance, for example acceleration and the metabolic cost of exercise [10]. Although only one study in the review reported body fat percentage using DXA [54], the gold standard for measuring body composition, all methods reported similar body fat % across age groups suggesting body composition remains reasonably stable. This was similar to cross sectional findings in RL although only skinfold thickness was reported [25]. Interestingly, longitudinal research in RL [90,91] and elite RU [92] has identified improvements in body composition during training periods, however no research has been carried out in age grade RU players to support this. Although similar results are reported for age grade players and elite international players (backs 10.7% and forwards 14.2%) [93], career progression from U16 to international as opposed to other playing levels was differentiated by a lower body fat percentage [47]. The acceptable body composition of age grade RU players is unknown, however further research into the longitudinal change and the interaction with other physical qualities is important for optimising both long and short-term performance.
Although there are limited observed differences between age grades, body composition is different between positions. Tight five and more specifically front row players are identified to have greater body fat than other positions [32,34]. Similar to height and body mass these are likely aligned with match demands with forwards playing fewer minutes, covering less distance and involved in more collisions [11,17].
Muscular strength. Muscular strength is important for RU performance due to the collision component of the sport and its relationship with other physical qualities (i.e. power and speed) [94]. This review found absolute muscular strength to be greater in older age grades. The differences observed between age grades is thought to be due to a combination of resistance training experience, and growth and maturation [60]. Similar to AFL [24] and RL [25], there is a paucity of research regarding the strength of age grade RU players, specifically players under the age of 16 years for measurements of lower body and upper body pull muscular strength. This may be a result of low training age where appropriate movement patterns are not yet developed for intense loads [95]. Future research should utilise testing methods which require less technique and therefore a lower injury risk while providing a valid and reliable measure of muscular strength such as the isometric mid-thigh pull previously used by Darrall-Jones et al. [40].
Forwards demonstrate greater muscular strength at all age grades compared to backs. It is favourable for forwards to possess greater absolute strength due to the positional demands placed on forwards performing more collisions, static high intensity efforts and lifting in the lineout [3]. Due to the greater emphasis on sprinting and explosive performance for the backs, relative strength data should be incorporated as the relationship between force production capabilities and body mass is important for optimising performance. Although increasing relative strength is important, its assessment in age grade RU is limited [38,40,60] with no positional comparisons. To further understand the positional differences of age grade RU players both absolute and relative strength measure should be reported.
Muscular power. Muscular power is important for success in collisions and the contact element of RU [10]. Similar to changes in height and mass, greater differences at younger age groups (� U15) were observed compared to older age groups (U16-U20) for both VJ and CMJ. At older age groups evidence utilising a cross sectional design is conflicting with Darrall-Jones et al. [40] observing differences in CMJ height between U18 and U18 while Kobal et al. [64] observed no difference between U17 and U19. The inconsistency observed in this review is similar to the AFL pathway and is suggested to be a result of the differences in maturity levels [24]. In contrast longitudinal research found increases in jump height during the playing season for U16 [60] and over multiple seasons from U18 to U20 [68]. Longitudinal research may provide a better explanation for the development of muscular power compared to cross sectional research.
Regarding playing position, backs generally have a greater jump height. It should be noted that although greater heights were reported for backs only Spammer & Hattingh [49] identified positional differences of practical significance at U18 and U19. Forwards jump performance may be constrained due to greater body mass and body fat percentage influencing their ability to exert force rapidly. Ball et al. [30] were the only study to use peak power to differentiate between positions with forwards producing greater power outputs. CMJ height may therefore be a poor proxy of muscular power when quantifying positional differences. There is no research to date which reports both direct and indirect measures of power for positional groups to suggest which method is superior. The use of both direct and indirect measures of power may be more appropriate to evaluate and monitor player development compared to jump height alone [40,41].
Linear speed. Linear speed is an important physical quality as it is associated with meters made, evasion and line and tackle breaks in senior RU [96]. Additionally, superior speed performance at U16 is suggested to differentiate career progression at the highest level [47]. This review suggests linear speed improves in younger age grades until U16 where further improvements become unclear. Kobal et al. [64] identified differences in 10, 20 and 30m sprint performance between U15 and older U17 and U19 players. No further increases in performance were found in studies that compared players competing at U16 and older [32,39,40]. As a result of growth, the development of longer limbs is suggested to influence stride length and frequency enhancing performance at a young age (< 16 years) before entering the period of peak weight velocity (> 16 years) which limits the development of speed qualities [80]. The constraint of body mass on speed development is supported by Casserly et al. [68] who suggest small increases in body mass act as a mediator of speed development from U18-U20. Furthermore, Barr et al. [71] identified sprint performance does not improve from U20 international to senior international, but rather body mass increases resulting in greater momentum. Both studies are however limited to age groups which are post maturation and only utilise a single team which may not reflect the variety of training interventions used within RU. Further longitudinal research is required to understand the concurrent development of body mass and speed during the earlier stages of growth and maturation.
Backs are shown to be quicker than forwards across a range of age groups and distances [32,36,39,49,57]. This is associated with the positional demands of RU where forwards are exposed to lower running demands while backs are often in open space with greater opportunities to run [97]. Although the differences between backs and forwards are consistent throughout the literature, the limited differences observed between individual positions may be exaggerated by the level of accuracy used by Durant [32], the only author to quantify linear speed performance of individual positions. Potentially due to the large amount of cross-sectional research utilising a small number of teams, positional differences are limited throughout this review as a result of small sample sizes. Future research should attempt to increase sample sizes using multiple clubs to increase the knowledge of positional differences in age grade RU players. Furthermore, due to a greater body mass compared to backs, forwards achieve a greater momentum even though they are observed to be slower [39]. In addition to the monitoring of speed qualities, changes in body mass should also be considered when goal setting and evaluating the development of positional speed qualities.
Change of direction. The ability to rapidly change direction is important in RU for match actions such as evasive running when attacking [98]. Findings suggest there are some differences in younger players U13-U16 [34,64], however differences in older players (> U16) are unclear with similar scores observed between age groups [32,34,40,64]. This contrasts findings in RL literature where a general improvement was observed with age, however the availability of change of direction data for older age grade RU players is limited [25]. These differences may be a result of the development of longer limbs during growth prior to an increase in body mass [80]. Increased body mass is thought to negatively effect change of direction ability as it increases the eccentric breaking required to reduce momentum and change direction [99]. This may explain an improvement in 505 results from U18 to U21 which are accompanied by increases in relative strength [40]. While the literature provides an insight of the change of direction ability of age grade RU players, no research was included within this review which assess the agility of RU players, where agility incorporates the response to a stimulus [100]. Future research should not only consider of the development and trainability of the components making up change of direction (i.e. accelerating, decelerating and reaccelerating), but also the ability to perform these actions in response to a stimulus.
Positional differences are less apparent in the research conducted. For both the Illinois agility test and T-test differences were unclear, with the exception of props [32,34]. This is in alignment with the findings in RL, but unlike RL, not all RU players are exposed to regular changes of direction due to defensive retreats [25]. It would therefore be beneficial to gain further understanding of positional specific movement signatures to assist with the development of change of direction ability.
Aerobic capacity. A well-developed aerobic capacity is needed for RU to be able to recover between high intensity bouts [19]. Current research is contradictory suggesting aerobic capacity does [32,39] and does not [39,40,64,68,73] differentiate between age grades. The contrast in findings may be due to the differences in tests used with Darrall-Jones et al. [39] identifying U18's cover a likely greater distance than U16 in the YYIR1, but no difference is observed for the 30-15IFT. Furthermore, when body mass is considered as a covariate differences are observed between age grades for both the 30-15IFT [73] and YYIR1 [68]. These findings suggest increases in body mass may mask expected improvements in aerobic capacity due to training and maturation during adolescents [101]. In contrast to these findings longitudinal research in rugby league has identified seasonal [90,102] and annual [85] improvements in aerobic capacity. Casserly et al. [68] provide the only study to longitudinally monitor the development of aerobic capacity in RU observing no change, however the age groups included were older than the RL players observed and post the effects of maturation. Utilising cross sectional data to assess the differences between age groups may not be appropriate and further longitudinal research may provide further information on the development of aerobic capacity and its relationship with body mass.
Positional differences are observed in the aerobic capacity of backs and forwards [32,36,39,57]. Forwards, more specifically props, are identified to be the worst performers [32]. Similar to linear speed and change of direction ability a greater body mass may influence performance in aerobic tests, especially those that contain a change of direction. The reduced aerobic capacity is associated with the playing demands of forwards who cover less ground and often have a reduced playing time compared to backs [103].

Limitations
First, a major limitation of the current literature is the diversity of tests used to quantify the physical qualities of age grade RU players. Although some common variables are identified throughout the literature, the lack of homogeneity in testing methods limits the effectiveness of the research when making comparisons between age grades across the review. Secondly, the research is limited by the large number of studies utilising a cross sectional design to compare between age grades. Although comparisons between independent samples provide a snapshot of differences between age grades, inferences cannot be made about the development of physical qualities on an individual level. Thirdly, the research regarding positional specific physical qualities is limited, potentially due to limited sample sizes. It is well documented that positional match demands vary greatly and therefore each will possess a specific set of physical qualities, however only two studies [32,34] report the physical qualities for positional sub groups (e.g., props, hookers, locks or tight forwards) as opposed to a unit (forwards and backs) or full team. Fourthly, there is a paucity of information regarding RU players under the age of 16 years. Understanding physical qualities below the age of 16 could enhance the prescription of physical development to align with an appropriate long-term athlete development programme. Finally, a limitation of this review was the inability to carry out a meta-analysis on the reported data providing summary normative values for the results presented. Due to the variety of tests, methods, positional groups and playing standards reported, no further analysis was carried out.

Future research
Future studies on the physical qualities of age grade RU players should build upon the current literature by increasing the availability of data using similar physical tests. It may be beneficial for researchers to work alongside practitioners or national governing bodies to develop national standardised testing batteries that are both practical and evidence based. The development of standardised testing batteries could provide the opportunity for studies to recruit samples from multiple clubs, thus increasing samples sizes, generalisability of results and statistical power of subcategory comparisons (e.g. position or playing level). It would also be advantageous to carry out longitudinal research on age grade RU players to better inform the development of physical qualities rather than the differences between two samples. The resultant normative values for changes in physical qualities would enhance the ability of practitioners to set achievable short and long-term goals. Furthermore, statistical modelling should consider the interactions between physical qualities (e.g. body mass and linear speed) and factors which may influence the development of physical qualities (i.e. growth and maturation), which have been discussed as potentially confounding factors within this review, to enhance the ability of practitioners to prescribe holistic training programs in accordance with long term athlete development models. Finally, further research should utilise innovative analysis techniques to increase the understanding of player selection and career progression throughout the talent identification pathway. Developments in data visualisation techniques should accompany this to increase the practicality of the data to all coaches and not just those with a statistical background.

Conclusion
Provision of normative data for the physical qualities of age grade RU players is important for practitioners to evaluate athlete performance, guide training prescription and inform goal setting. This is the first systematic review to collate the tests used throughout the literature to identify the physical qualities of age grade RU players and present the current evidence by age grade and position. Seventy-five tests were identified to assess seven physical qualities (body composition, muscular strength, muscular power, linear speed, change of direction ability, aerobic capacity and anaerobic endurance capacity). When comparing the physical qualities between age grades differences are apparent between younger age grades (� U16). Although older age groups (U19-U20) generally performed the best in physical testing, increased physical performance was not always clear between U16 and U20 age grades, except for muscular strength. The differences at all age groups are potentially due to factors such as increased training exposure and growth and maturation. Relative measures of physical qualities could further distinguish between age grades to account for increases in body mass associated with training, growth and maturation. Positional differences observed are often related to match demands with forwards being taller, heavier and stronger while backs are faster and fitter and therefore training should reflect these differences. The normative data presented in this review can be used by practitioners to evaluate the physical qualities of age grade rugby players and subsequently prescribe appropriate training programmes. The practical use of the data is limited however by the variety of testing methods used, lack of positional data and the paucity of longitudinal research. The use of standardised testing batteries may be beneficial for further research to guide the physical development of age grade RU players.