REVIEW ARTICLE

Abstract
Water polo has been played for over a century. While the rules of the game have evolved considerably over this time the sport has consistently remained, physiologically, a highly demanding activity. Much attention has been paid to the technical and strategic elements of the game; however, despite the potential for improvements in athletic performance and the maintenance of athletes’ health, there are few published studies (particularly in English) on the physical and physiological demands and adaptations to water polo training and competition.

Game analyses have demonstrated that water polo is an intermittent sport comprised of intense bursts of activity of <15 seconds duration with intervening, lower intensity intervals averaging <20 seconds duration. Physiological measurements obtained during game play show a cumulative effect of the repeated sequences of activities and suggest there is a high metabolic demand on the athletes. The multiple individual skills and movements required for playing water polo also place considerable demands on the neuromuscular system. Observations of the frequency and duration of the different activities, and of the physiological responses to participating in a water polo match, are initial sources of information for designing training programs specific to the game and to the different playing positions.

The physical and physiological attributes of elite water polo players offer insight into the minimum requirements for participation and the adaptations that result from training and competition. Further systematic documentation and experimentation are required to facilitate the design and specification of individual training programs and to better understand the long term effects of water polo on athletes’ health.

1. The Game of Water Polo
Water polo originated in the UK in the late 1800s. The sport had its Olympic Games debut in Paris in 1900, making it the earliest, and longest-running, team competition in the modern Olympic Games. Historically, participation has been dominated by the European nations, but the game has developed into a popular activity in North America, Australia and Asia. With the recent inclusion of women’s water polo as an event at the Sydney Olympic Gams in the year 2000, further international participation and development of the sport is anticipated.

International water polo games currently consist of 4 x 7-minute periods of stop-time play separated by 2-minute intervals. In the event of a tie in deciding rounds. 2 x 3-minute periods, with a 1-minute interval, are played 5 minutes after the end of regulation time. In the event of a continued tie, single 3-minute periods of ‘sudden death’ are played. The duration of international and national championship games, excluding any overtime, is approximately 55 minutes (32 to 60 minutes). Each quarter lasts about 12 minutes and the actual running time between quarters averages 2.5 minutes. When extra time is required, game duration can extend to >70 minutes.

The game is intermittent in nature, requiring a variety of different, intense bursts of activity, each lasting <15 seconds. The high intensity activities are interspersed by others of lower intensity, and similar duration. and by occasional longer periods of recovery, such as the intervals between game quarters. There are very few restrictions on the movement of players during the frequent stoppages in the actual playing time, so players are rarely inactive at any time during each quarter of a game.

The specialized playing positions of water polo are the goaltenders and the field players, who include athletes who play primarily in the center ward, center defense or wing positions. During a game, field players will frequently rotate among and share, these primary positions. Of the 13 players on a team, usually 2 will be goaltenders and the remainder field players. The division of the 11 field players varies between teams but might include center forwards, 3 center defenders and 5 players whose primary responsibility is in the wing position. A maximum of 1 goaltender and 6 field players are in the water at any time.

Despite a large number of coaching publications on the technical and tactical aspects of water polo, research-based literature on the game is limited. Earlier studies focused on the morphological characteristics of players, aspects of drag and propulsion during water polo activities and the biomechanics of the ‘egg-beater’ kick, front crawl and overhead throwing techniques. Only more recently have the physiological characteristics of water polo players, and the physiological demands of the sport and its training practices, been evaluated.

This article reviews the documented physical and physiological demands of the game and the characteristics of the athletes. as played at senior national and international levels. Available information for both male and female athletes is included, as is data from physiological laboratory and sport-specific field testing of the athletes. Applications for sport-specific training and areas for further physiological research in the sport are then suggested.

2. Physical and Physiological Demands of Water Polo
As with most team sports, the physical and further physiological demands of water polo depend not only on the rules of the game but also on the characteristics of each particular game. The nature of the demands of water polo, as determined or influenced by the rules, is presented in table 1. Characteristics specific to an individual game that may affect the demands of play are also included.

International rules are reviewed biannually following Olympic Games and World Championship competitions, and changes are implemented after every World Championships. Since 1982. several rule modifications have been introduced which may have affected the physiological demands of the game. These include increasing the number of players on a team from 11 to 13, increasing the length of each quarter from 5 to 7 minutes of stoptime play and shortening the maximum duration of exclusions to 20 seconds. In addition, there are more opportunities for player substitutions, and coaches may initiate a limited number of (1-minute) time-outs. Goaltenders may now throw the ball anywhere within the field of play and goals may be scored from the awarding of a free-throw following minor fouls occurring at, or beyond, 7 meters from the goal line.

In general, the purpose of these rulings was to make the game faster and more spectacular. The actual effect of these rule changes on the nature of the game, from a physiological perspective, is unknown. Similarly, despite the international rule differences for males and females (e.g. the field of play is larger for men, and women use a smaller ball), the physiological demands of the game as played by women have not been adequately addressed.

Few studies on the actual demands of water polo have been published. Given the potential variation in these demands, between games and among athletes, more systematic observation of the game under different sets of conditions is warranted. It is also important that the physical demands of particular games, and of individual players can be manipulated by players and coaches. This is primarily accomplished by the alteration of the recovery intervals or by rotation of the type, intensity and frequency of particular activities during play, without necessarily compromising the overall intensity level of all activities performed by a team. Such manipulations have the potential to increase collective physical performance of an elite team over match and tournament play. The effects of manipulations such as player substitutions, player combinations in the water (positions played in rotation and the calling of time-outs on physiological demands and responses (i.e. fatigue and performance) of players have not been considered in published studies.

2.1 General Demands of Water Polo
The physical activities performed while playing water polo are numerous and can be categorized according to the skills involved, the rules which delineate the activities and the velocity of movement. The duration and frequency of the performance of each activity by individual players, often supplemented by indications of intensity such as velocity or displacement, can be assessed by video analyses. An overview of the average duration of all periods of activity, and rest intervals, observed during games at a men’s National Championship are presented in table II. The work: rest ratios and active time as a percentage of the quarter and of total game times are also shown.

The complexity, variation and intermittent nature of water polo make the assessment and interpretation of the actual physiological responses to water polo training and competition technically difficult. Physiological measurements, such as heart rate monitoring and/or capillary blood lactate analyses have provided indications of the cumulative cardiovascular and metabolic demands of the identified activities during games. Data from video analyses and the reported physiological measurements are combined in the following sections to present a picture of the demand of field play and goaltending. Very little information on the neuromuscular demands of the game is available, yet biomechanical observations of throwing and shooting support the empirical observations that water polo players require moderately high levels of muscular strength and power.

2.2 Demands of Field Play
A summary of the intensity, duration, and frequency of the multiple activities performed by players during competitive men’s games is shown in table III. Video analyses have shown that activities in games of water polo last for <20 seconds, with intense movements and sprints averaging only 7 to 14 seconds. The singular performance of an activity of this intensity and brevity is likely to be highly dependent on anaerobic metabolism and muscular power. However, these activities are in fact performed in sequences, resulting in cumulative, and hence longer duration, moderate and high intensity periods of exertion which constitute approximately two-thirds of total game time. Heart rate monitoring during play has demonstrated that athletes spend a similar proportion of game time with heart rates which, during tethered swimming, correspond to intensities >80 maximum aerobic power (VO 2 max). For almost all of the time players were in the water, their rates exceeded 80% of maximum, suggesting that the intervening lower intensity, activities were of insufficient duration for recovery, ensuring a persistent, moderately high demand on aerobic metabolism.

Observations of players during international competition have shown that for approximately 85% of the time, the velocities of movement in horizontal plane were below that which, during constant-velocity free swimming, elicited blood lactate levels of <2 mmol/L. These data have been interpreted as reflecting a high demand from the anaerobic lactic system, a high demand on the aerobic system for the replenishment of creatine phosphate and a lesser emphasis on anaerobic lactic metabolism for energy provision. However, glycolytic metabolism is greatly accelerated at the onset of a maximal intensity. In addition, blood samples taken during play, and following game quarters, have shown individual lactate levels ranging from 2 to 12 mmol/L, and mean values of 7 to 9 mmol/L, in elite Spanish and Italian male players. Individual blood lactate levels ranging from 2 to 10 mmol/L, and mean values of 5 mmul/L following game quarters, have also been demonstrated in Dutch national level female players.

In combination with the high heart rates recorded during games, such blood lactate values indicate a moderate demand on the anaerobic glycolytic (lactic) system. It is probable that the velocity of horizontal displacement does not adequately reflect the intensity and intermittent nature of the activities performed, particularly for acceleration and deceleration, movements in the vertical plane or on contact with an opponent.

For athletes to rely upon, and tolerate, the increased flux through anaerobic metabolism may be undesirable during competition. yet there appears to be occasion for a moderate demand on anaerobic glycolysis during play. This energy system may also be important during training sessions, and should be specifically considered in the planning of training programs. Furthermore, the half-time for lactate removal following maximal swimming exercise has been shown to be approximately 12 to 20 minutes, and varies with the intensity of activity during the recovery interval. This suggests that players who accumulate high levels of lactate early in a game will have insufficient opportunity to fully recover, and this may impede their performance during subsequent play.

Swimming for the central purpose of transportation comprises 20% of the total game time (11 to 13 minutes) and about 33% of the time a player is active in the water. About one-half of this time is spent sprinting, i.e. near maximal velocity. Over the course of a game, a player swims for transport an average of 60 times, each occasion lasting 10 to 12 seconds (13 to 15 meters). Swimming movements performed to gain the advantage over an opponent, or to defend against attacking maneuvers, are of similar durations. Because these activities are difficult to quantify in terms of linear displacement, they are not included in linear swimming distances estimated but are included as separate actions (see table III). Difficulties in assessing the distance swum, and the apparent large degree of variation in these estimates, can be accounted for by the number of brief swimming movements which are more rectilinear and often include opponent contact. Nevertheless, calculations of the total linear distance traveled by male players during an entire game range from an average of 500 to 1000 meters up to 1500 to 1800 meters.

Intuitively. the distance traveled will vary considerably amongst players and between games depending upon the nature of the particular game, the style of play, the positions played and the technical abilities of individual players. However, for applications to competition-specific training, calculations from time and motion analyses indicate that field players spend only 45 to 55% of game time in a horizontal body position (i.e. as in transportational swimming). The remainder of the time is spent performing activities in predominantly vertical body positions, with and without contact with an opponent, and at a moderate to high intensity as indicated by heart rate recordings. In terms of physiological stress, these activities include the requisite change from a vertical to a horizontal body position. With this transition there is the initiation of swimming and the acceleration from 0 to >2 meters/second on 60 or mom occasions per game.

There are positional differences in the skills and abilities required of field players. Center forwards spend more time in contact with opponents, spend longer in a vertical body position, have more frequent transitions between horizontal and vertical body positions and perform a different variety of actions at different frequencies than other field players. In contrast, center forward players may not perform as many brief (<10 seconds) attacks or overhead shots as players in the wing positions. Although coaches apply position-specific conditioning exercises, the physiological implications of playing in the center forward, center defense or wing positions have not been presented. A systematic identification of the positional demands of the game would have implications for the establishment of specific (in-water and on-land) training programs for individual athletes.

2.3 Demands of Goaltending
Video tape analyses of the durations and frequencies of the activities of male goaltenders in water polo have demonstrated a consistency in the demands of different athletes, and over the course of 10 different games (table IV). All activities classified as being medium to high intensity were of brief duration (<15 seconds). Brief, explosive movements were infrequent, but were often required of athletes following 10 to 15 seconds of moderately intense work. The activities occurred in sequences lasting approximately 35 seconds. Following each series of activities there was a predictable interval of easy sculling which averaged 47 seconds in duration. and thus allowed for some recovery. Mean blood lactate levels (5 to 6 mmol/L) measured in elite male goaltenders following each quarter of Spanish National League water polo games indicate a moderate cumulative effect of the multiple brief activities performed.

The mean frequency and duration of each activity performed during the game did not vary significantly between quarters, nor among the individual goaltenders. On this basis, the efficacy of a senior competitive goaltender may depend upon the quality of the execution of particular skills (e.g. timing, velocity, height, angle and/or body position) rather than how often, or for how long, the activities are performed.

2.4 Neuromuscular Demands
Water polo players spend considerable time and effort training out of the pool to increase muscular strength and power for performance. and in the interest of injury prevention. The actual neuromuscular demands of water polo and the contribution of muscular strength, power and endurance to aspects of performance are not well understood. Several biomechanical studies have quantified the relative neuromuscular demands of individual water polo activities. Electromyographic studies demonstrating a near-continual static activity of several muscle groups during throwing, and the involvement of numerous muscle groups during swimming, suggest relatively high levels of activity and force generation during the multiple movements involved in water polo.

The measured ball velocities at the release of overhand throws in competitive male water polo players have ranged from 16.5 to 19.9 meters/second. Documented velocities at release during shooting in 6 elite female players averaged 14.7 meters/second. In throwing, the ball is accelerated from 0 meters/second to these peak velocities in less than one-quarter of a second. Considering the size and mass of a water polo ball, and that the athletes lack a solid base of support and must overcome the resistance of the water, these throws require considerable muscular power. As a comparison, the maximal throwing velocities are quite similar to those reported for overhand throwing in American football.

The neuromuscular demands of goaltending have not been quantified in the literature. The documented activities of goaltenders suggest a greater dependence on brief, intense muscular movement of a different nature to those

most often required of field players. Goaltending activities predominantly include specific types of jumps and the maintenance of a high vertical body position with both arms extended overhead. More sustained activity includes vertical sculling of moderate high intensities. In addition, goaltenders rarely shoot the ball, but make frequent long (10 to 20 meters) passes during games and practices.

A quantitative indicator of the impact forces a goaltender might counteract when blocking a shot can be derived from force plate measurements. Recorded impact forces of overhand shots from 4 and 8 meters away, by Belgian national level players. ranged from 402 to 981N, although ball contact time was <0.01 seconds. Shot-blocking techniques often aim to manipulate counterforces produced by the goaltenders during the ball contact phase, presumably increasing ball contact time and reducing or redistributing these high impact forces. Nonetheless, goaltenders require high muscular power and the necessary strength and joint stability of the upper body to cope with repeated ball impact.

2.5 Demands of Training and Tournament Play
In addition to examining the physiological demands of water polo games, the demands of, and responses to repeated training sessions and tournament play should be considered. Current practices for elite teams include active training sessions totaling up to 4 to 6 hours per day for 6 days a week. Some national tournaments include up to 2 or 3 games per day and run for 3 to 4 days. International tournaments involve a maximum of 1 game per team per day, but. usually, 8 to 10 games are played over a 10- to 12-day period. The schedules will have some influence upon the responses of the athletes to training and competition and both the competitive tactics and training programs must take this into consideration. The effects of training, and responses over the duration of tournament play, have not been addressed in the literature.

2.6 Adaptations to Water Polo Training
Physiological adaptations to playing water polo over the course of training cycles, seasons or years are not well documented, particularly by longitudinal studies (see Hohmann for examples). Performance and anthropometric indices for developing players have been proposed, yet there is little evidence of the magnitude, or complete nature, of long term training-induced alterations in physiological function in adult water polo players. One cross-sectional study reported that the spinal and hip bone density was higher in male water polo players than in nonexercising controls, and similar between water polo players and active, weight training males. The higher bone density in both exercising groups was attributed to their increased habitual physical activity and muscular strength, independent of the type of the exercise. Importantly, this may illustrate a benefit from water polo training on bone density, despite the non-weight-bearing nature of most of the exercise performed. Compared with sedentary controls, water polo players have also been known to have higher plasma levels of high-density lipoproteins and higher levels of lipid metabolic enzymes. which are of importance in protecting against atherosclerosis and may be positively related to habitual physical exercise.

Seasonal adaptations in water polo players, assessed predominantly by performance measures, have been infrequently presented and have concentrated on a few indices of metabolic and/or cardiovascular function. Recent advances in understanding endocrine, immune and neuromuscular function and adaptation in response to different modes of exercise and training have been reviewed in the general sports science literature, and highlight the potential for a better appreciation of how such systems influence, and are influenced by, the combination and specificity of training modes used by water polo players.

3. Physical Characteristics of Water Polo Players
The physical characteristics of elite water polo players reflect the various and physically intense demands of the sport, and can be used by coaches in guiding athletes to specialize in playing in the various positions. A summary of the physical characteristics of male and female water polo players is shown in Table V. A full description of the anthropometric characteristics of a large sample of players competing in the 1991 World Championships in Perth, Australia has recently been published. Readers are referred to this publication for further detail.

Potential changes in body composition with water polo training and the relevance of an individual’s body size and composition to water polo performance have not been well studied. Analysis of team data from the 1991 World Championships and the 1996 Olympic Games (unpublished observations) indicate that at the top international level, the mean physical size of the players on a team have no bearing on that team’s results. This may be due to both the varied player body size within teams and to the evaluation of success by the outcome of the team rather than individual players. However, a comparison of water polo players at the 1995 World Student (FISU) Games held in Fukuoka, Japan has indicated that the top 8 teams had significantly greater average stature and body mass than the teams placed in the bottom half of the competition (unpublished observations). Player eligibility and team entry criteria for this international tournament resulted in a wider variation in the teams’ standards of performance than at the Olympic Games or World Championships. This may account for the descrepancy with the observations noted above.

Positional differences in player body size are apparent between goaltenders and field players. There is also variation among the different field positions, even though players often fill different positions and roles during and between games. Athletes at the 1991 World Championships who played primarily in the center forward or center defence positions were taller, heavier and dimensionally larger (breadths, lengths, girths) than the wing players. Goaltenders were also taller and heavier and had longer upper and lower limb lengths and a greater arm span that the wing players. Male goaltenders were lighter, and female goaltenders of similar body mass, compared with their center forward and center defence counterparts. Two exceptional field players participating in the 1996 Olympic Games exemplify the range in body size possible for elite play. One athlete was 1.78 meters tall and weighed 82 kilograms, and the other was 2.01 meters tall and weighed 108 kilograms.

Indicators of body composition (sum of skinfold thickness measurements, total body muscle mass and total body skeletal mass) show no differences between female players playing in the different positions. Male players were similarly homogeneous in body composition with respect to playing position, the only exception being the goaltenders who had a greater relative skeletal mass than did the center forwards.

Presumably, these and the anthropometric differences are a result of both natural selection and coaching decisions, and reflect suitability for the physical demands of the position. I suggest that there are minimal size atributes required for top international play. Such attributes are likely to be specific to each playing position and, for any given individual, may depend upon their technical skill and the size characteristics of the other team members.

4. Physiological Characteristics of Water Polo Players
Data describing the physiological capabilities of water polo players can reflect the demands of training and game performance. Several studies have provided an indication of the physiological characteristics of male players, yet little data on female players have been published. Similarly, data describing the capacities of athletes according to playing position are limited. Laboratory testing of players has been minimal. Field tests have the advantage of increased specificity, applicability and potential relevance to game performance; however, the number of different tests used by different coaches, and the lack of an established test validity or reliability, has hindered their consistent use and interpretation. Tests using objective physiological measures - often based upon laboratory-like established test protocols - have been used, and may serve to form a basis for a valid and reproducible sport-specific test battery. Performance test data, used to describe the physiological capabilities of water polo players, is included in sections 4.1 to 4.7.

4.1 Aerobic Power and Endurance
From the large proportion of moderate and high intensity activity, and the sustained elevated heart rates observed during games, aerobic metabolism probably provides the majority of energy required for playing water polo. The VO2max values from national and international-level field players, obtained by direct methods during free or tethered swimming, range from 4.5 to 4.7 L/min. or 58 to 61 ml/kg/min. These moderately high levels of aerobic power are similar to those reported for players of other intermittent contact team sports such as rugby, ice hockey and basketball. Data from top international teams or athletes (e.g.Olympic Games participants) and from female players are lacking.

Direct measurements of VO2max (4.2 L/min. 58 ml./kg/min. (unpublished observations) during treadmill running in 1 international-level female water polo player, over a 4 year period, suggest that a moderately high aerobic power is also a characteristic of elite female players. As with elite competitive swimmers, VO2max may not be greatly altered (e.g. < 5%) over a training season, and any changes in this measure may not be adequately indicated by treadmill or cycle ergometer testing. The relative importance of aerobic power to game performance, particularly amongst international-level competitors, has not been established. There may be a minimum aerobic power requirement for international-level play, above which differences among athletes may not contribute greatly to game performance.

Maximal aerobic power has been determined indirectly in competitive swimmers but, to date, no swimming test has been established to predict maximal aerobic power in water polo players. Because of the variation in the oxygen cost of swimming between players (unpublished observations), a predictive equation may have a large standard error of the estimate for these athletes.

Repeated swims of increasing velocity up to a maximal effort with post-swim blood lactate measurements and continuous heart rate monitoring have been used to monitor training effectiveness in water polo players. Distances ranging from 200 to 400 meters for each submaximal swim, and from 100 to 400 meters for the maximal swim, have been used. Comparisons of blood lactate levels and mean heart rate, plotted as a function of mean swimming velocity between repeated tests on individual athletes, can represent adaptations in aerobic metabolism. For the assessment of changes in aerobic endurance, a minimum of 2 x 300-meter submaximal swims and a 1 x 400 meter maximal effort is recommended for senior field players. Adequate rest must be allowed between swims and blood samples should be taken at consistent intervals following each swim (e.g. 3 minutes post-swim). Waterproof heart rate monitors are now widely available and the receiver can easily be placed under the athlete’s bathing cap to record values during each swim.

4.2 Swimming Economy
Swimming economy, or the relationship between VO2 and swimming velocity, has been used as a practical measure to reflect the technical efficiency of an athlete in forward swimming locomotion. Water polo players are less economical and have greater interindividual variation in economy during head-down, linear, front-crawl swimming compared with competitive swimmers. At velocities ranging from 1.1 to 1.5 meters per second, elite male water polo players have approximately 6 to 20% greater oxygen consumption than competitive swimmers (unpublished observations). Such differences may be a function of stroke technique and/or the physical characteristics of the athletes which affect hydrodynamic drag or propulsion.

The relevance of economy, as measured during continuous head-down, front-crawl swimming, to water polo performance may be limited, given that players are almost always accelerating or decelerating in terms of their horizontal movement during games. In addition, for over one-half of playing time, various different activities are performed in vertical or lateral directions, often in combination, and almost always with the head above the surface of the water. However, the effectiveness of propulsive technique in achieving these skills may contribute to their level of execution. As during front-crawl swimming, an effective technique may reduce the overall energy cost of play and the accumulated fatigue over a game. Nevertheless, the assessment of oxygen cost, heart rate and blood lactate level, over a range of swimming velocities, can be used as an indicator of training efficacy and adaptation, particularly in developing players.

4.3 Anaerobic Power and Capacity
The short duration and intensity of individual activities and the blood lactate levels that can be accumulated during play suggest that anaerobic power and capacity are important to performance. Data from laboratory testing of water polo players’ anaerobic power and capacity have been reported, although their relevance to game performance, or to game-specific anaerobic capabilities, is difficult to assess. Maximal 30-second arm crank ergometry tests in 28 Canadian National Team players, using a resistance of 0.7 g/kg body mass, showed peak 5-second (alactic) power of 497W and mean power (anaerobic capacity) of 353W (10 590J) (unpublished observations). A mean power output of approximately 485W (29 118J) during a 60-second maximal effort on a cycle ergometer by 17 Hungarian water polo players has also been reported.

More sport-specific indicators of anaerobic capacity have included measurement of peak capillary blood lactate levels following maximal 100-meter swims or 60-second swimming ergometer tests. Mean blood lactate values of 13 to 16 mmol/L were observed in international-caliber male players following such tests, and these reflect the ability of these athletes to accumulate moderately high levels of lactate following a single maximal effort.

4.4 Muscle Strength
The specific muscular strength of water polo players has been documented primarily by isokinetic dynamometric measurements of peak torque generation by the shoulder. To my knowledge, there are no published reports of muscular strength, power or endurance of female water polo players. In male players, indices of neuromuscular function have always been assessed out of the pool and have been primarily aimed at the detection of antagonist-agonist torque imbalances and their association with joint injury, or in relation to overhead shooting performance.

As isokinetic measurements are specific to the equipment and protocol used, values should be considered in terms of bilateral and agonist-pair ratios, or on a test-retest basis, within the same athlete. Data from American and Canadian (unpublished observations) national team members have shown the internal rotation torque of the shoulder to be greater in the dominant arm. Ratios of external: internal rotation were not significantly different between the arms nor among the 3 different positions tested. However, these ratios are lower in water polo players than in nonathletic controls. A low external: internal rotation torque ratio has been implicated as a contributing factor to various shoulder injuries observed in water polo players and other overhead throwing athletes, and in swimmers. Prospective studies of the relationships between shoulder torque imbalances, shooting and swimming techniques, and injuries in water polo players are warranted. The influence of strengthening programs aimed at maintaining, or attaining, minimum shoulder antagonist : agonist torque ratios on the incidence of injury, and in the rehabilitation from injury, should also be examined.

The use of muscular strength, endurance or power measurements in monitoring of training effectiveness, or in relation to game performance, has rarely been considered. Isometric grip strength and forearm extension and rotational strength measured by cable tensiometry have shown low correlations (r = 0.5) with maximal overhead shooting velocity. An 8-week isotonic strength training program, superimposed on the in-water training, increased arm rotation strength above that of non-strength trained controls, yet this measure showed a low correlation with overhead shooting velocity. As with strength measurements obtained using isokinetic dynamometry, the specificity of both training and testing modes, and the muscle groups, range of motions, contraction type, velocity of movement, etc., should be considered when evaluating the relevance of strength measures to training effectiveness or water polo skill performance.

4.5 Flexibility
Flexibility, defined as the range of motion about a specific joint or series of joints, has seldom been assessed in water polo players. Coaching practices and the training habits of the athletes support a belief in the importance of flexibility to aspects of water polo performance. Stretching exercise and its associated temporary or persistent increase in flexibility are believed to contribute to a reduced chance of injury and/or improved skill performance. Theoretically, an increased range of motion allows for an increased distance through which an active force can be applied, with a greater resultant force or torque. This would be applicable to movements such as overhead throws or vertical jumps using an ‘eggbeater’ -type kick.

Simple static trunk flexion in male water polo players indicated that they were no more flexible in this movement than either field hockey or rugby players. Neither trunk flexion, trunk extension nor shoulder extension measures in water polo players have significantly correlated with overhead shot velocity. However, the specificity of the range of motion to each particular joint and movement pattern involved, as well as the influence of an individual’s antropometrics on several measures, limits the interpretation and application of flexibility data. In addition, a wide interindividual variation in the range of motion of shoulder rotation and flexion-extension (coefficient of variation 11 to 13%) has been observed, even among members of the same national team (unpublished observations). No significant differences between the preferred shooting arm and the nondominant arm were apparent in these measures. The relevance of these and other flexibility measures to performance or to the incidence of musculoskeletal injury in water polo players has yet to be determined.