Sports Med 2009; 39 (7): 513-522 0112-1642/09/0007-0513/$49.95/0
LEADING ARTICLE
ª 2009 Adis Data Information BV. All rights reserved.
Youth Sports in the Heat Recovery and Scheduling Considerations for Tournament Play Michael F. Bergeron National Institute for Athletic Health and Performance and Center for Youth Sports and Health, Sanford USD Medical Center, Sioux Falls, South Dakota, USA
Abstract
One of the biggest challenges facing numerous young athletes is attempting to perform safely and effectively in the heat. An even greater performance challenge and risk for incurring exertional heat injury is encountered when a young athlete has to compete multiple times on the same day, with only a short rest period between rounds of play, during a hot-weather tournament. Within the scope of the rules, tournament directors frequently provide athletes with only the minimum allowable time between same-day matches or games. Notably, prior same-day exercise has been shown to increase cardiovascular and thermal strain and perception of effort in subsequent activity bouts, and the extent of earlier exercise-heat exposure can affect performance and competition outcome. Incurred water and other nutrient deficits are often too great to offset during short recovery periods between competition bouts, and the athletes are sometimes ‘forced’ to compete again not sufficiently replenished. Providing longer rest periods between matches and games can significantly improve athlete safety and performance, by enhancing recovery and minimizing the ‘carryover’ effects from previous competitionrelated physical activity and heat exposure that can negatively affect performance and safety. Governing bodies of youth sports need to address this issue and provide more specific, appropriate and evidence-based guidelines for minimum rest periods between same-day contests for all levels of tournament play in the heat. Youth athletes are capable of tolerating the heat and performing reasonably well and safely in a range of hot environments if they prepare well, manage hydration sufficiently, and are provided the opportunity to recover adequately between contests.
Youth sports provide myriad physical and social health-enhancing benefits, improvements in fitness, and enjoyment for participating young athletes.[1-3] Measurable gains in cardiorespiratory health and capacity, motor skills, and general and functional muscular strength, endurance and power, as well as better body composition and bone mineral content and density, have been shown in boys and girls as a result of regular
participation in organized youth sports.[4-16] With increased participation, training and competition, however, there is also a greater risk of injury,[1,17-23] and for those sporting events held in hot, humid environments, a particular concern to athletes, parents, coaches and medical support staff is the potential for adverse effects on cardiovascular and thermal strain, performance, and exertional heat injury risk.[24-40] The physical
Bergeron
514
challenges are even greater when children and adolescents attempt to perform safely and effectively through multiple strenuous competition bouts on the same day during a hot-weather tournament. This is a common scenario for numerous young athletes involved in organized youth sports – soccer and tennis tournaments are notable examples. Unfortunately, reasonable safety and optimal performance are often compromised by short rest periods between same-day contests during tournament play in the heat. This article addresses the physiological challenges in youth sports during tournament format competitions in the heat, where multiple contests (games or matches) played on the same day can put young athletes at greater risk for poor performance and exertional heat injury, seemingly due, in large part, to insufficient rest and recovery time between competition bouts. 1. Same-Day Repeated Bouts During tournament play in organized youth sports, young athletes often must compete two or more times on the same day. Within the scope of the rules, tournament directors frequently provide athletes with only the minimum allowable time between same-day matches or games. Empirical observations from parents and coaches indicate that, with only a short recovery period between contests, a young athlete generally has considerably less tolerance to the subsequent activity and often succumbs to the strain of the cumulative physiological demands by way of poor performance or withdrawal from play. This is particularly obvious in hot environments. 1.1 Rest and Recovery Period Guidelines
Depending on the sport, level and age-group of event, number of team or individual entries, and other logistical and administrative considerations, the amount of time between competitions in organized youth sports tournaments can vary considerably. With certain sports such as softball and baseball, tournament rules have a particular focus on pitching limitations (number of pitches and innings per day and subsequent ª 2009 Adis Data Information BV. All rights reserved.
rest days), although most governing associations also provide recommendations, based on age, for limiting the number of games that can be played per day or throughout a tournament. Minimum time between same-day contests for these and many other organized youth sports, however, is not typically defined in the tournament playing rules. Table I shows examples of youth tournament (or meet) single-day scheduling from actual events at the divisional or regional level. It is particularly worth noting the minimum rest periods between tournament games and matches in youth sports such as soccer and tennis, given the extensive physical demand on these young athletes throughout each round of play.[41,42] In the US, the duration of youth soccer games (e.g. 25to 45-minute halves) and overtime periods are based on age, and the prescribed minimum rest period between tournament games can typically range from 1 to 2 hours depending on the sanctioning association, although local tournament events often provide much less time. In junior tennis, minimum rest periods between tournament matches are defined by the national governing body. Again, in the US, where a young player can have up to five matches scheduled in one day (singles and doubles combined), the minimum rest period between matches in the junior divisions is 1 hour (30 minutes for shorter pro set formats),[43] with no consideration for the duration of each previous same-day match or Table I. Examples of youth tournament (or meet) single-day scheduling from actual events at the divisional or regional level. Recommendation: longer recovery periods (minimum two times longer than shown here), especially in hot environmental conditions and/or following a long match or game Sport
Competition load
Basketball (indoor) 3–6 games/day
Between-contests rest 1 game duration (or less)
Soccer
2–3 games/day
1–2 h or 1–2 time slots
Softball
2–3 games/day
15 min (minimum)
Baseball
2–5 games/day
15 min–1 h
Track and field
3–5 running events/day
1 h (or less)
Tennisa
3 singles, 2 doubles 1 h (singles), 30 min matches/day (doubles)
a
No time limit for each match; a singles match can last, for example, less than 1 h or up to 4 h or more.
Sports Med 2009; 39 (7)
Youth Sports Scheduling in the Heat
level of environmental heat stress. In any junior division in tennis, a match can take less than 1 hour to complete or go on for up to 4 hours or more. Notably, public school systems generally have very specific (albeit quite variable between systems) guidelines for scheduling and suspending practice and competition in extreme heat. In contrast, such restrictions are typically not as specifically outlined for or consistently utilized in organized youth sports tournaments, and it is often left up to the discretion of the on-site tournament administrator or referee to adjust or alter the competition format based on the environment. National-level and scholastic tournament events also typically provide more conservative guidelines to minimize the number of same-day contests, compared with regional, state and local non-scholastic events where tournament directors and referees often do not follow the same recommendations and practices. Unfortunately, the potential negative impact of previous competition-related physical activity and heat exposure on subsequent same-day performance and exertional heat injury risk in children and adolescents participating in organized youth sports is not well described or sufficiently appreciated. However, related field and laboratory studies on repeated-bout exercise in adults and those very few similar new studies on young athletes can provide some valuable insight into these potential after- or ‘carryover’ effects during tournament play in youth sports. 1.2 Adult Studies
Examining the effects of environmental heat stress during successive days of Marine Corps basic training, Wallace et al.[44] found that exertional heat illness risk increased with hourly wetbulb-globe temperature (WBGT; 11% increase in risk per F [~0.5C]). Although limiting physical activity during hot weather measurably reduced exertional heat illness risk, there was still an elevated relative risk and 17% of exertional heat illness cases occurring across the cautionary flag conditions for training activities. Moreover, exertional heat illness risk was associated with the ª 2009 Adis Data Information BV. All rights reserved.
515
current WBGT as well as with the previous day’s average WBGT – that is, they observed a strong cumulative effect on exertional heat illness risk resulting from current environmental conditions and certain exercise-heat strain effects carried over from the day before. According to Wallace et al., those who schedule training or sports activities should consider the carryover effects of previous-day exercise and heat stress exposure to better anticipate heat-related problems and minimize exertional heat injury risk. These data further suggest that exertional heat illness risk should still be considered on a day that appears ‘safe’, if the previous day’s heat stress was high. In contrast, McLellan et al.[45] reported no carryover effects from previous-day moderate exercise-heat exposure on core body temperature or perceived exertion in sedentary non-heatacclimated (but well hydrated) adults during similar exercise on the next morning. Notably, the previous day’s afternoon exercise bout lasted only 1 hour and was followed by a single morning bout in less hot conditions. Despite these factors, a number of subjects failed to complete the exercise trial in the heat on day 2. Ronsen et al.[46] examined the residual effects of prior same-day exercise (cycling for 75 minutes) and varying recovery times on metabolic responses during a subsequent bout of the same exercise in elite endurance-trained athletes. They found augmented metabolic stress and greater cardiovascular and thermal strain (higher heart rate and core body temperature, respectively) in the second exercise bout with 3 hours between bouts, which were significantly attenuated by providing a longer rest and recovery period (6 hours) between the exercise sessions. These data underscore how the residual effects of prior exercise can affect one’s response to a subsequent same-day exercise session (even without a hot environment) and how the extent of these carryover effects can be modulated by increasing the length of the rest and recovery time between exercise bouts. Several other relevant studies provide some additional insight into same-day repeated-bout exercise. Sawka et al.[47] observed a higher core body temperature response and greater cardiovascular Sports Med 2009; 39 (7)
Bergeron
516
strain (elevated heart rate and lower stroke volume and cardiac output) in trained runners during a second of two 80-minute bouts of highintensity treadmill running at room temperature, separated by a 90-minute rest period. Kruk et al.[48] examined consecutive moderate exercise bouts (cycling for 30 minutes) with 30-minute rest periods in between and found progressively higher core body temperature and heart rate responses with each successive exercise period, despite replacement of sweat fluid loss and no environmental heat stress. Similarly, in examining the cumulative effect of prior cycling exercise for 30 minutes in a hot environment on another identical exercise bout following a 45-minute post-exercise recovery period of seated rest in the same heat, Brenner et al.[49] reported greater rectal temperature and cardiovascular responses during the second session, as well as greater perceived exertion. Lastly, Yamada and Golding[50] also observed greater rectal temperature and heart rate during a second exercise bout (walking for 23 minutes) in the heat, following rehydration and rest during a short (12-minute) recovery period after the first identical bout of exercise. Notably, the subjects in these studies were either insufficiently or inappropriately hydrated during the test sessions and/or core body temperature had not returned to baseline between exercise bouts. These findings are particularly relevant to certain youth sports (e.g. soccer and tennis) in also highlighting how a 10-minute break between halves or sets (or a little longer between games or matches) is insufficient time to sufficiently lower rectal temperature and avert incurring greater physiological strain when activity in the heat is resumed, even with rehydration, especially if the players stay in the heat during the break. 1.3 Youth Studies
To test the hypothesis that young athletes would experience an increase in physiological strain and related ratings of perceived discomfort during a subsequent identical exercise bout (compared with the first), even when maintaining adequate hydration, Bergeron et al.[51] examined 24 healthy, young athletes (12–13 and 16–17 years ª 2009 Adis Data Information BV. All rights reserved.
old) during two 80-minute intermittent exercise (treadmill and cycle ergometer) sessions in the heat (33C), with a 1-hour rest and recovery period (21C) between bouts. Core body temperature, physiological strain index[52] and perceived thermal stress were similar during the second exercise bout. However, a 1-hour recovery period was not sufficient to avert greater perception of effort and, for some, greater physiological strain during the second bout of identical exercise, even when the young athletes consumed ample fluid during exercise, drank enough between the exercise bouts to offset any remaining fluid deficit, and core body temperature returned to baseline prior to starting exercise again. It is important to recognize that there were still apparent carryover effects to the second bout of exercise, even in ‘ideal’ recovery conditions – i.e. immediate rest and no heat exposure, full rehydration, and complete cool-down. Such circumstances are in contrast to more typical ‘real life’ tournament scenarios, where incomplete thermoregulatory recovery and subsequent increased thermal strain[48,49] are more likely to occur, because the players cannot immediately begin and maintain complete rest and cool-down procedures during the recovery period, and the next round of play may be contested in more stressful environmental conditions a little later in the day. Coyle[53] examined data from a national boys 14s (14-year-olds) junior tennis championships event over a 7-year period to show how cumulative heat stress can affect a second (same-day) match outcome during tournament play. With the effect of seeding removed, the winner of an afternoon singles match could be effectively predicted from the same-day degree minutes (product of WBGT in C and length of match [including warm-up] in minutes) acquired during the morning matches by the two respective players. Coyle found no statistical differences in hometown heat stress zone between match winners and losers, although all of these young athletes had been training and competing in the heat for some time prior to this event and thus were likely heat-acclimatized and able to tolerate the heat better than if they began play without sufficient recent exercise-heat exposure.[38,54-56] Sports Med 2009; 39 (7)
Youth Sports Scheduling in the Heat
From the same national boys 14s event, Bergeron et al.[28] also showed how junior tennis players tend to not fully rehydrate from earlier sameday play and how even just doubles competition in the afternoon following singles play can elicit an appreciable thermal challenge in a hot environment. 2. Nutrient Recovery Needs and Challenges Fluid and electrolyte deficits and carbohydrate needs can be substantial following a long and intense contest in the heat. It might be argued that those young athletes who incur a substantial fluid deficit and perform poorly or develop heatrelated problems in subsequent tournament rounds simply are not rehydrating sufficiently. Accordingly, from this perspective, the focus should be on education and compliance versus adjusting the schedule. Indeed, as with adults, a child’s or adolescent’s hydration status has a direct impact on cardiovascular and thermal strain and heat tolerance during exercise and sports.[28,38,39,57-59] With sweat rates potentially ranging from 300 to 700 mL/h in 9- to 12-year-olds[38,39,51,60] and 1–2 L or more per hour in older adolescents,[28,29,51] it is not surprising that young athletes can incur significant water and electrolyte losses from sweating during competition.[61] Much of this can be compensated by appropriate fluid intake during play, if opportunities to rehydrate are sufficiently frequent. However, in certain sports (e.g. soccer), a very active and continuously engaged participant may have few to no opportunities to consume fluid during the game (except at breaks between periods). The process of offsetting a post-play body water deficit can usually begin right away; however, there is a practical limit (often short of sufficiently eliminating the deficit) to the amount of fluid a young athlete can comfortably and safely consume with only a short recovery period.[62] For example, a moderate sweating rate of only 1.5 L/h could lead to a ‡5 L total sweat fluid loss for an adolescent player over the course of a long tennis match. Even with drinking regularly throughout the match, this young player could readily be facing a ª 2009 Adis Data Information BV. All rights reserved.
517
>2 L fluid deficit by the end of play. Full rehydration (»2–3 L)[63] before the next match begins may not be possible if the time between matches is only an hour or so. If the cycle continues over the course of several matches on the same day, the player could potentially reach a point where the fluid deficit is too great to perform well or safely. Although fluid intake and hydration status are not always statistically associated with core body temperature when young players have the opportunity and desire to reduce exercise intensity and effort,[29] during a meaningful and intense tournament competition in the heat, the negative effects of insufficient hydration on thermal strain can be more readily apparent.[28] Notably, during and after play, effective rehydration requires more than just ample fluid intake. Sodium losses need to be replaced as well, so that the ingested fluid is more optimally retained and distributed to all fluid compartments, resulting in more complete rehydration.[63-66] Given the extensive sweat losses and deficits of fluid and electrolytes and energy (carbohydrate) depletion potentially incurred during play, the nutrient intake, absorption and distribution required to replace these deficits may sometimes be too great to achieve or tolerate in the allotted recovery time.[67-72] Thus, the young athlete is ‘forced’ to begin play again not optimally or sufficiently replenished. Incomplete rehydration and a sodium deficit can prompt lower heat tolerance, greater cardiovascular and thermal strain, and reduced performance,[73-77] as well as an increased risk for developing muscle cramps[61,78] during the next game or match. Moreover, low glycogen stores following a previous exercise bout and short recovery period[79] can prompt a greater pro-inflammatory cytokine response,[80] further increase the inherent challenge to maintain euglycaemia,[81] and prompt earlier fatigue[82-87] during the next round of play.
3. Who Should Assume Responsibility? The responsibility for preparing to compete in the heat should be assumed by the athlete, parent(s) or primary care provider(s), and coach.[88] Sports Med 2009; 39 (7)
Bergeron
518
Young athletes also have the responsibility to rehydrate adequately when there is sufficient time to offset a post-play fluid deficit. This is not always the case, and players often begin competition not well hydrated when there was ample opportunity to do so;[28,89] albeit, being well hydrated prior to beginning play does not guarantee minimum thermal strain or exertional heat illness risk. However, tournament directors and other event administrators, referees, medical support personnel, and ultimately the various sport governing bodies, have the responsibility to provide safe and appropriate venues and scheduling in youth sports.[88] Some might argue that the body of published scientific and clinical evidence is not yet conclusive and therefore any changes to the current scheduling rules and practices are premature. However, an alternative perspective is emphasized by Emery et al.,[88] who point to a broadened interpretation of the precautionary principle as a guide for injury prevention in youth sports: ‘‘When there is suspected harm and the scientific evidence is inconclusive, the prescribed course is precautionary action’’ – i.e. to side with safety and reduce the risk.[90] This seems particularly appropriate when considering potential harm to children and adolescents. As indicated earlier, there is growing evidence in the literature strongly supporting the perspective that insufficient rest and recovery time, especially when competing in an already unsafe climate, can further reduce performance and put certain young athletes at increased clinical risk, even if they are fit, willing to play and motivated to win. Although desirable, these attributes are no match for a schedule that is inappropriate and does not emphasize athlete safety over completing multiple rounds of competition quickly.
4. Specific Research Needs Previous studies claim that children have a disadvantage compared with adults during exercise in the heat because boys and girls are less effective in regulating body temperature, and thus are less tolerant to and capable of performing well in a hot environment.[57,59,91-94] However, ª 2009 Adis Data Information BV. All rights reserved.
recent research and more appropriate comparisons do not indicate that children (9–12 years old) have less effective thermoregulation, insufficient cardiovascular capacity or lower tolerance during exercise in the heat when environmental and exercise conditions, aerobic fitness and heat acclimatization status are relatively the same for the children and adults and, importantly, hydration is maintained.[38,39,60] Considerably more research specific to exercise-heat tolerance in children needs to be done, however, before any definitive conclusions can be made to defend or dismiss the notion of inherently greater cardiovascular or thermal strain or other clinical risk in children or adolescents compared with adults during sports competition in the heat, especially with multiple same-day contests in a tournament format. Prospectively examining the effects of different recovery approaches (including utilizing various rehydration strategies and cooling methods[95] and durations between bouts of competition, based on environmental conditions) on subsequent physiological strain and other outcome measures related to athlete performance and safety is essential. This would provide important insight to minimizing the potential negative carryover effects of previous competition-related physical activity and heat stress exposure on clinical risk and performance during subsequent same-day competition bouts. Comprehensive epidemiological studies[96,97] on exertional heat injury and performance during outdoor tournament play is also a priority. Such research is critical in being able to provide the most appropriate evidencebased guidelines to youth sports tournament directors, administrators and sport governing bodies for safely and effectively scheduling multiple competition bouts in the heat. This will help to optimize performance, while reducing the risk of exertional heat injury, for numerous young athletes who regularly participate in a number of different organized youth sports. It is important to also consider other influential factors from a previous bout of demanding competition-related physical activity that affect athlete functional capacity and safety. Especially if recovery time between matches or games Sports Med 2009; 39 (7)
Youth Sports Scheduling in the Heat
519
Table II. Key scheduling challenges and related effects on young athletes, and recommended responses for event administration, that can improve athlete safety and performance in the heat Challenges
Effects
Recommended responses
Hot and humid environment Multiple same-day contests Long matches or games (i.e. extensive exercise-heat exposure)
Increased physiological strain Reduced activity tolerance and performance Incomplete recovery between competition bouts Greater risk for exertional heat injury
Increase recovery time between same-day contests Minimize number of contests per day for each athlete Monitor athletes more closely Provide verbal reminders regarding hydration and cooling strategies Schedule contests during cooler times of day Consider cancelling the event altogether, if environmental conditions are extreme
is short, muscle and central fatigue and acute muscle injury/soreness can independently or in combination measurably affect sensorimotor acuity, neuromuscular control, joint stability, and even temperature regulation[95,98-106] and thus potentially decrease performance and increase injury risk during the next same-day contest. These aspects should be comprehensively addressed as well in organized youth sports to fully appreciate the clinical and performance impact of too much same-day competition and insufficient rest and recovery between rounds of play. 5. Key Points and Recommendations1 The key points and recommendations below should be considered by youth sports event administrators and others when developing policies and event schedules, so as to enhance the health and safety of the participating athletes. Key challenges and effects, and recommended responses for event administration, that can improve athlete safety and performance in the heat are also highlighted in table II. Multiple matches or games on the same day can pose particular performance and safety challenges to young athletes and increase exertional heat injury risk, due to insufficient recovery time and rehydration, as well as physiological ‘carryover’ effects from previous- and sameday competition-related physical activity and heat exposure.
Youth sports tournament directors, administrators and governing bodies should recognize and appreciate that providing longer rest and recovery periods between matches and games (much more than many current governing bodies’ rules indicate) as environmental heat stress increases can improve athlete safety and performance without necessarily having to adjust the playing format (e.g. scoring or duration of the individual contests). Following a long and intense match or game in the heat, a young athlete may have substantial water, electrolyte, carbohydrate and other macronutrient recovery needs. This requires a proportionately longer rest and recovery period, so that s/he has the opportunity to begin the next contest without one or more significant nutrient deficits or a full stomach and consequent performance and safety disadvantages. Recovery time between contests is not the only significant scheduling-related concern when running a tournament in hot weather. Competing at all in an unsafe environment can put young athletes at great risk, even if they are well rested and sufficiently hydrated before play. In conditions of uncompensable heat stress (where evaporative cooling is insufficient to maintain thermal balance), player safety should be the priority, and contests should be cancelled or rescheduled to cooler times, even if it means playing very early or later in the evening under the lights.
1 Repeated same-day training sessions in the heat also present similar safety and performance risks; accordingly, the recovery and scheduling considerations for tournament competition and the recommendations presented here should be considered and applied to youth sports training and practice in the heat as well.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Bergeron
520
Acknowledgements No sources of funding were used to assist in the preparation of this article. The author has no conflicts of interest that are directly relevant to the content of this article.
References 1. Committee on Sports Medicine and Fitness and Committee on School Health. Organized sports for children and preadolescents. Pediatrics 2001; 107: 1459-62 2. Sports and children: consensus statement on organized sports for children. FIMS/WHO Ad Hoc Committee on Sports and Children. Bull World Health Organ 1998; 76: 445-7 3. Strong WB, Malina RM, Blimke CJR, et al. Evidence based physical activity for school-age youth. J Pediatr 2005; 146: 732-7 4. Bencke J, Damsgaard R, Saekmose A, et al. Anaerobic power and muscle strength characteristics of 11 years old elite and non-elite boys and girls from gymnastics, team handball, tennis and swimming. Scan J Med Sci Sports 2002; 12: 171-8 5. Duncan CS, Blimkie CJ, Cowell CT, et al. Bone mineral density in adolescent female athletes: relationship to exercise type and muscle strength. Med Sci Sports Exerc 2002; 34: 286-94 6. Dyson K, Blimkie CJ, Davison KS, et al. Gymnastic training and bone density in pre-adolescent females. Med Sci Sports Exerc 1997; 29: 443-50 7. Filaire E, Lac G. Nutritional status and body composition of juvenile elite female gymnasts. J Sports Med Phys Fitness 2002; 42: 65-70 8. Ginty F, Rennie KL, Mills L, et al. Positive, site-specific associations between bone mineral status, fitness, and time spent at high-impact activities in 16- to 18-year-old boys. Bone 2005; 36: 101-10 9. Hansen L, Bangsbo J, Twisk J, et al. Development of muscle strength in relation to training level and testosterone in young male soccer players. J Appl Physiol 1999; 87: 1141-7 10. Kawabe H, Murata K, Shibata H, et al. Participation in school sports clubs and related effects on cardiovascular risk factors in young males. Hypertens Res 2000; 23: 227-32 11. Laing EM, Massoni JA, Nickols-Richardson SM, et al. A prospective study of bone mass and body composition in female adolescent gymnasts. J Pediatr 2002; 141: 211-6 12. Laing EM, Wilson AR, Modlesky CM, et al. Initial years of recreational artistic gymnastics training improves lumbar spine bone mineral accrual in 4- to 8-year-old females. J Bone Miner Res 2005; 20: 509-19 13. Leone M, Lariviere G, Comtois AS. Discriminant analysis of anthropometric and biomotor variables among elite adolescent female athletes in four sports. J Sports Sci 2002; 20: 443-9 14. McMillan K, Helgerud J, Macdonald R, et al. Physiological adaptations to soccer specific endurance training in professional youth soccer players. Br J Sports Med 2005; 39: 273-7
ª 2009 Adis Data Information BV. All rights reserved.
15. Roetert EP, Piorkowski PA, Woods RB, et al. Establishing percentiles for junior tennis players based on physical fitness testing results. Clin Sports Med 1995; 14: 1-21 16. Tsunawake N, Tahara Y, Moji K, et al. Body composition and physical fitness of female volleyball and basketball players of the Japan inter-high school championship teams. J Physiol Anthropol Appl Human Sci 2003; 22: 195-201 17. American Academy of Pediatrics Committee on Sports Medicine and Fitness. Injuries in youth soccer: a subject review. Pediatrics 2000; 105: 659-61 18. American Academy of Pediatrics Committee on Sports Medicine and Fitness. Risk of injury from baseball and softball in children. Pediatrics 2001; 107: 782-4 19. Brenner JS. Overuse injuries, overtraining, and burnout in child and adolescent athletes. Pediatrics 2007; 119: 1242-5 20. Pickett W, Streight S, Simpson K, et al. Head injuries in youth soccer players presenting to the emergency department [commentary]. Br J Sports Med 2005; 39: 226-31 21. Pluim BM, Staal JB, Windler GE, et al. Tennis injuries: occurrence, aetiology, and prevention. Br J Sports Med 2006; 40: 415-23 22. Radelet MA, Lephart SM, Rubinstein EN, et al. Survey of the injury rate for children in community sports. Pediatrics 2002; 110: 28-38 23. Spinks AB, McClure RJ. Quantifying the risk of sports injury: a systematic review of activity-specific rates for children under 16 years of age. Br J Sports Med 2007; 41: 548-57; discussion 557 24. Allman Jr FL. The effects of heat on the athlete. J Med Assoc Ga 1992; 81: 307-10 25. American Academy of Pediatrics Committee on Sports Medicine and Fitness. Climatic heat stress and the exercising child and adolescent. Pediatrics 2000; 106: 158-9 26. Armstrong LE, Maresh CM. Exercise-heat tolerance of children and adolescents. Pediatr Exerc Sci 1995; 7: 239-52 27. Bergeron MF, McKeag DB, Casa DJ, et al. Youth football: heat stress and injury risk. Med Sci Sports Exerc 2005; 37: 1421-30 28. Bergeron MF, McLeod KS, Coyle JF. Core body temperature during competition in the heat: National Boys’ 14’s Junior Tennis Championships. Br J Sports Med 2007; 41: 779-83 29. Bergeron MF, Waller JL, Marinik EL. Voluntary fluid intake and core temperature responses in adolescent tennis players: sports beverage versus water. Br J Sports Med 2006; 40: 406-10 30. Binkley HM, Beckett J, Casa DJ, et al. National Athletic Trainers’ Association position statement: exertional heat illnesses. J Athl Training 2002; 37: 329-43 31. Bytomski JR, Squire DL. Heat illness in children. Curr Sports Med Rep 2003; 2: 320-4 32. Casa DJ, Almquist J, Anderson S, et al. Inter-Association Task Force on Exertional Heat Illnesses consensus statement. NATA News 2003; Jun: 24-9 33. Eichner E. Heat stroke in sports: causes, prevention, and treatment. Sports Sci Exchange 2002; 15: 1-4 34. Gonzalez-Alonso J, Mora-Rodriguez R, Below PR, et al. Dehydration markedly impairs cardiovascular function in
Sports Med 2009; 39 (7)
Youth Sports Scheduling in the Heat
35.
36.
37. 38.
39.
40.
41.
42. 43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
hyperthermic endurance athletes during exercise. J Appl Physiol 1997; 82: 1229-36 Gonza´lez-Alonso J, Teller C, Andersen SL, et al. Influence of body temperature on the development of fatigue during prolonged exercise in the heat. J Appl Physiol 1999; 86: 1032-9 Gutierrez G. Solar injury and heat illness: treatment and prevention in children. Physician Sportsmed 1995; 23: 43-8 Hoffman JL. Heat-related illness in children. Clin Pediatr Emerg Med 2001; 2: 203-10 Rivera-Brown AM, Rowland TW, Ramirez-Marrero FA, et al. Exercise tolerance in a hot and humid climate in heat-acclimatized girls and women. Int J Sports Med 2006; 27: 943-50 Rowland T, Garrison A, Pober D. Determinants of endurance exercise capacity in the heat in prepubertal boys. Int J Sports Med 2007; 28: 26-32 Sawka MN, Burke LM, Eichner ER, et al. American College of Sports Medicine position stand: exercise and fluid replacement. Med Sci Sports Exerc 2007; 39: 377-90 Fernandez J, Mendez-Villanueva A, Pluim BM. Intensity of tennis match play. Br J Sports Med 2006; 40: 387-91; discussion 391 Stølen T, Chamari K, Castagna C, et al. Physiology of soccer: an update. Sports Med 2005; 35: 501-36 Barber B, Cummings J, Farley B, et al. Friend at court: the USTA handbook of tennis rules and regulations. 2008 ed. White Plains (NY): US Tennis Association Inc., 2008 Wallace RF, Kriebel D, Punnett L, et al. The effects of continuous hot weather training on risk of exertional heat illness. Med Sci Sports Exerc 2005; 37: 84-90 McLellan TM, Rhind SG, Bell DG. Body temperature in sedentary adults during moderate exercise: no effect from exercise the day before. Aviat Space Environ Med 2002; 73: 1167-75 Ronsen O, Haugen O, Hallen J, et al. Residual effects of prior exercise and recovery on subsequent exerciseinduced metabolic responses. Eur J Appl Physiol 2004; 92: 498-507 Sawka MN, Knowlton RG, Critz JB. Thermal and circulatory responses to repeated bouts of prolonged running. Med Sci Sports 1979; 11: 177-80 Kruk B, Szczypaczewska M, Opaszowski B, et al. Thermoregulatory and metabolic responses to repeated bouts of prolonged cycle-ergometer exercise in man. Acta Physiol Pol 1990; 41: 22-31 Brenner IK, Zamecnik J, Shek PN, et al. The impact of heat exposure and repeated exercise on circulating stress hormones. Eur J Appl Physiol Occup Physiol 1997; 76: 445-54 Yamada PM, Golding LA. Physiological responses of female fire fighters to repeated work bouts. AMAA J 2004; 17: 5-7 Bergeron MF, Laird MD, Marinik EL, et al. Repeatedbout exercise in the heat in young athletes: physiological strain and perceptual responses. J Appl Physiol 2009; 106: 476-85 Moran DS, Shitzer A, Pandolf KB. A physiological strain index to evaluate heat stress. Am J Physiol 1998; 275: R129-34
ª 2009 Adis Data Information BV. All rights reserved.
521
53. Coyle J. Cumulative heat stress appears to affect match outcome in a junior tennis championship [abstract]. Med Sci Sports Exerc 2006; 38: S110 54. Rowland T, Hagenbuch S, Pober D, et al. Exercise tolerance and thermoregulatory responses during cycling in boys and men. Med Sci Sports Exerc 2008; 40: 282-7 55. Shapiro Y, Moran D, Epstein Y. Acclimatization strategies: preparing for exercise in the heat. Int J Sports Med 1998; 19 Suppl. 2: S161-3 56. Sunderland C, Morris JG, Nevill ME. A heat acclimation protocol for team sports. Br J Sports Med 2008; 42: 327-33 57. Bar-Or O, Dotan R, Inbar O, et al. Voluntary hypohydration in 10- to 12-year-old boys. J Appl Physiol 1980; 48: 104-8 58. Dougherty KA, Baker LB, Chow M, et al. Two percent dehydration impairs and six percent carbohydrate drink improves boys basketball skills. Med Sci Sports Exerc 2006; 38: 1650-8 59. Falk B, Bar-Or O, MacDougall JD. Thermoregulatory responses of pre-, mid-, and late-pubertal boys to exercise in dry heat. Med Sci Sports Exerc 1992; 24: 688-94 60. Inbar O, Morris N, Epstein Y, et al. Comparison of thermoregulatory responses to exercise in dry heat among prepubertal boys, young adults and older males. Exp Physiol 2004; 89: 691-700 61. Bergeron MF. Heat cramps: fluid and electrolyte challenges during tennis in the heat. J Sci Med Sport 2003; 6: 19-27 62. Gisolfi CV. Is the GI system built for exercise? News Physiol Sci 2000; 15: 114-9 63. Shirreffs SM, Maughan RJ. Volume repletion after exercise-induced volume depletion in humans: replacement of water and sodium losses. Am J Physiol 1998; 274: F868-75 64. Mitchell JB, Phillips MD, Mercer SP, et al. Postexercise rehydration: effect of Na+ and volume on restoration of fluid spaces and cardiovascular function. J Appl Physiol 2000; 89: 1302-9 65. Sanders B, Noakes TD, Dennis SC. Sodium replacement and fluid shifts during prolonged exercise in humans. Eur J Appl Physiol 2001; 84: 419-25 66. Sanders B, Noakes TD, Dennis SC. Water and electrolyte shifts with partial fluid replacement during exercise. Eur J Appl Physiol Occup Physiol 1999; 80: 318-23 67. Burke LM. Nutrition for post-exercise recovery. Aust J Sci Med Sport 1997; 29: 3-10 68. Burke LM, Kiens B, Ivy JL. Carbohydrates and fat for training and recovery. J Sports Sci 2004; 22: 15-30 69. Goetze O, Steingoetter A, Menne D, et al. The effect of macronutrients on gastric volume responses and gastric emptying in humans: a magnetic resonance imaging study. Am J Physiol Gastrointest Liver Physiol 2007; 292: G11-7 70. Ivy JL. Muscle glycogen synthesis before and after exercise. Sports Med 1991; 11: 6-19 71. Jentjens RLPG, Wagenmakers AJM, Jeukendrup AE. Heat stress increases muscle glycogen use but reduces the oxidation of ingested carbohydrates during exercise. J Appl Physiol 2002; 92: 1562-72
Sports Med 2009; 39 (7)
522
72. Shirreffs SM, Taylor AJ, Leiper JB, et al. Post-exercise rehydration in man: effects of volume consumed and drink sodium content. Med Sci Sports Exerc 1996; 28: 1260-71 73. Cheung SS, McLellan TM. Heat acclimation, aerobic fitness, and hydration effects on tolerance during uncompensable heat stress. J Appl Physiol 1998; 84: 1731-9 74. Montain SJ, Coyle EF. Influence of graded dehydration on hyperthermia and cardiovascular drift during exercise. J Appl Physiol 1992; 73: 1340-50 75. Moran DS, Montain SJ, Pandolf KB. Evaluation of different levels of hydration using a new physiological strain index. Am J Physiol 1998; 275: R854-60 76. Sawka MN. Physiological consequences of hypohydration: exercise performance and thermoregulation. Med Sci Sports Exerc 1992; 24: 657-70 77. Sawka MN, Montain SJ, Latzka WA. Hydration effects on thermoregulation and performance in the heat. Comp Biochem Phys 2001; 128: 679-90 78. Bergeron MF. Heat cramps during tennis: a case report. Int J Sport Nutr 1996; 6: 62-8 79. Blom PC, Hostmark AT, Vaage O, et al. Effect of different post-exercise sugar diets on the rate of muscle glycogen synthesis. Med Sci Sports Exerc 1987; 19: 491-6 80. Ronsen O, Lea T, Bahr R, et al. Enhanced plasma IL-6 and IL-1ra responses to repeated versus single bouts of prolonged cycling in elite athletes. J Appl Physiol 2002; 92: 2547-53 81. Galassetti P, Mann S, Tate D, et al. Effect of morning exercise on counterregulatory responses to subsequent, afternoon exercise. J Appl Physiol 2001; 91: 91-9 82. Costill DL. Sweating: its composition and effects on body fluids. Ann NY Acad Sci 1977; 301: 160-74 83. Burke LM, Hawley JA. Fluid balance in team sports: guidelines for optimal practices. Sports Med 1997; 24: 38-54 84. Maughan R, Shirreffs S. Recovery from prolonged exercise: restoration of water and electrolyte balance. J Sport Sci 1997; 15: 297-303 85. Mohr M, Krustrup P, Bangsbo J. Fatigue in soccer: a brief review. J Sports Sci 2005; 23: 593-9 86. Rockwell MS, Rankin JW, Dixon H. Effects of muscle glycogen on performance of repeated sprints and mechanisms of fatigue. Int J Sport Nutr Exerc Metab 2003; 13: 1-4 87. Williams MB, Raven PB, Fogt DL, et al. Effects of recovery beverages on glycogen restoration and endurance exercise performance. J Strength Cond Res 2003; 17: 12-9 88. Emery CA, Hagel B, Morrongiello BA. Injury prevention in child and adolescent sport: whose responsibility is it? Clin J Sport Med 2006; 16: 514-21 89. Petrie HJ, Stover EA, Horswill CA. Nutritional concerns for the child and adolescent competitor. Nutrition 2004; 20: 620-31 90. Pless IB. Expanding the precautionary principle. Inj Prev 2003; 9: 1-2
ª 2009 Adis Data Information BV. All rights reserved.
Bergeron
91. Drinkwater B, Kupprat I, Denton J, et al. Response of prepubertal girls and college women to work in the heat. J Appl Physiol 1977; 43: 1046-53 92. Falk B. Effects of thermal stress during rest and exercise in the paediatric population. Sports Med 1998; 25: 221-40 93. Haymes EM, Buskirk ER, Hodgson JL, et al. Heat tolerance of exercising lean and heavy prepubertal girls. J Appl Physiol 1974; 36: 566-71 94. Wagner J, Robinson S, Tzankoff S, et al. Heat tolerance and acclimatization to work in the heat in relation to age. J Appl Physiol 1972; 33: 616-22 95. Barnett A. Using recovery modalities between training sessions in elite athletes: does it help? Sports Med 2006; 36: 781-96 96. Caine D, Caine C, Maffulli N. Incidence and distribution of pediatric sport-related injuries. Clin J Sport Med 2006; 16: 500-13 97. Goldberg AS, Moroz L, Smith A, et al. Injury surveillance in young athletes: a clinician’s guide to sports injury literature. Sports Med 2007; 37: 265-78 98. Benjaminse A, Habu A, Sell TC, et al. Fatigue alters lower extremity kinematics during a single-leg stop-jump task. Knee Surg Sports Traumatol Arthrosc 2008 Apr; 16 (4): 400-7 99. Cheung SS, Sleivert GG. Multiple triggers for hyperthermic fatigue and exhaustion. Exerc Sport Sci Rev 2004; 32: 100-6 100. Davey PR, Thorpe RD, Williams C. Fatigue decreases skilled tennis performance. J Sports Sci 2002; 20: 311-8 101. Girard O, Lattier G, Maffiuletti NA, et al. Neuromuscular fatigue during a prolonged intermittent exercise: application to tennis. J Electromyogr Kinesiol 2008 Dec; 18 (6): 1038-46 102. Girard O, Lattier G, Micallef JP, et al. Changes in exercise characteristics, maximal voluntary contraction, and explosive strength during prolonged tennis playing. Br J Sports Med 2006; 40: 521-6 103. Montain SJ, Latzka WA, Sawka MN. Impact of muscle injury and accompanying inflammatory response on thermoregulation during exercise in the heat. J Appl Physiol 2000; 89: 1123-30 104. Ronglan LT, Raastad T, Borgesen A. Neuromuscular fatigue and recovery in elite female handball players. Scand J Med Sci Sports 2006; 16: 267-73 105. Rozzi SL, Lephart SM, Fu FH. Effects of muscular fatigue on knee joint laxity and neuromuscular characteristics of male and female athletes. J Athl Train 1999; 34: 106-14 106. Tripp BL, Yochem EM. Uhl TL. Functional fatigue and upper extremity sensorimotor system acuity in baseball athletes. J Athl Train 2007; 42: 90-8
Correspondence: Prof. Michael F. Bergeron, National Institute for Athletic Health and Performance, Sanford USD Medical Center, 1210 W 18th Street, Suite 204, Sioux Falls, SD 57104, USA. E-mail:
[email protected]
Sports Med 2009; 39 (7)
Sports Med 2009; 39 (7): 523-546 0112-1642/09/0007-0523/$49.95/0
REVIEW ARTICLE
ª 2009 Adis Data Information BV. All rights reserved.
Medial Tibial Stress Syndrome A Critical Review Maarten H. Moen,1 Johannes L. Tol,2 Adam Weir,2 Miriam Steunebrink2 and Theodorus C. De Winter2 1 Department of Sports Medicine of the University Medical Centre Utrecht and Rijnland Hospital, Leiderdorp, the Netherlands 2 Department of Sports Medicine of the Medical Centre Haaglanden, the Hague, the Netherlands
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Literature Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Aetiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Functional Anatomy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Biomechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Histology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Patient Examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Physical Examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Radiograph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Bone Scan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.4 High-Resolution Computed Tomography Scan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.5 Imaging Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Risk Factor Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Risk Factor Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Conservative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Prevention Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Prevention Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract
523 525 525 525 525 532 532 533 533 533 534 534 534 535 536 537 537 537 538 538 538 539 540 540 541 541 543
Medial tibial stress syndrome (MTSS) is one of the most common leg injuries in athletes and soldiers. The incidence of MTSS is reported as being between 4% and 35% in military personnel and athletes. The name given to this condition refers to pain on the posteromedial tibial border during exercise, with pain on palpation of the tibia over a length of at least 5 cm. Histological studies fail to provide evidence that MTSS is caused by periostitis as a result of traction. It is caused by bony resorption that outpaces bone
Moen et al.
524
formation of the tibial cortex. Evidence for this overloaded adaptation of the cortex is found in several studies describing MTSS findings on bone scan, magnetic resonance imaging (MRI), high-resolution computed tomography (CT) scan and dual energy x-ray absorptiometry. The diagnosis is made based on physical examination, although only one study has been conducted on this subject. Additional imaging such as bone, CT and MRI scans has been well studied but is of limited value. The prevalence of abnormal findings in asymptomatic subjects means that results should be interpreted with caution. Excessive pronation of the foot while standing and female sex were found to be intrinsic risk factors in multiple prospective studies. Other intrinsic risk factors found in single prospective studies are higher body mass index, greater internal and external ranges of hip motion, and calf girth. Previous history of MTSS was shown to be an extrinsic risk factor. The treatment of MTSS has been examined in three randomized controlled studies. In these studies rest is equal to any intervention. The use of neoprene or semi-rigid orthotics may help prevent MTSS, as evidenced by two large prospective studies.
Medial tibial stress syndrome (MTSS) is one of the most common causes of exercise-induced leg pain.[1] Incidences varying from 4% to 35% are reported, with both extremes being derived from military studies.[2-4] This condition is most frequent among military personnel, runners and athletes involved in jumping, such as basketball players and rhythmic gymnasts.[5,6] There is much controversy about the definition and terminology of this condition. Different authors have used different names, such as ‘shin soreness’,[7] ‘tibial stress syndrome’,[8] ‘medial tibial syndrome’,[9] ‘medial tibial stress syndrome’,[10] ‘shin splints syndrome’[11] and ‘shin splints’.[12] In this review we chose to use ‘medial tibial stress syndrome’ because, in our opinion, this best reflects the aetiology of the syndrome. MTSS is characterized by exercise-related pain on the posteromedial side of the mid- to distal tibia. In 1966 the American Medical Association defined the condition (then termed shin splints) as: ‘‘pain or discomfort in the leg from repetitive running on hard surfaces or forcible excessive use of the foot flexors; diagnosis should be limited to musculotendinous inflammations, excluding fracture or ischaemic disorder.’’[13] This definition is the only available official definition given in the literature, but in our opinion is outdated and was never well accepted among clinicians. It does not ª 2009 Adis Data Information BV. All rights reserved.
describe signs on physical examination. Frequently when in the (older) literature the term ‘shin splints’ is used, ‘medial tibial stress syndrome’ is meant. More recently, an updated and better definition was proposed by Yates and White.[4] They described MTSS as ‘‘pain along the posteromedial border of the tibia that occurs during exercise, excluding pain from ischaemic origin or signs of stress fracture.’’ Additionally, they stated that on palpation with physical examination, a diffuse painful area over a length of at least 5 cm should be present. However, since no official definition exists, many authors use their own definition of MTSS. This makes comparison between studies difficult. Before diagnosing MTSS, the diagnosis of tibial stress fracture and exertional compartment syndrome should be excluded (see section 4). Detmer[14] in 1986 developed a classification system to subdivide MTSS into three types: (i) type I – tibial microfracture, bone stress reaction or cortical fracture; (ii) type II – periostalgia from chronic avulsion of the periosteum at the periosteal-fascial junction; and (iii) type III – chronic compartment syndrome. In the recent literature, stress fracture and compartment syndrome are qualified as separate entities. The objective of this review is to provide a critical analysis of the existing literature on MTSS. Aetiology, biomechanics, histology, patient Sports Med 2009; 39 (7)
Medial Tibial Stress Syndrome
evaluation, diagnostic imaging, risk factors, therapy and prevention are discussed.
525
Table I. Assessment of methodological quality and level of evidence (reproduced from Institute for Quality and Healthcare, the Netherlands,[15] with permission)
1. Methods
Assessment of methodological quality of studies concerning intervention (treatment/prevention)
1.1 Literature Search
A1: Systematic review of at least two independently conducted studies of A2 level
The electronic databases MEDLINE (1966–2009), EMBASE (1980–2009), CINAHL (1982–2009), SPORTDiscus (1975–2009) and Cochrane Library were searched for articles. The search terms ‘shin splints’, ‘medial tibial syndrome’, ‘medial tibial stress syndrome’ and ‘tibial stress syndrome’ were used with no restrictions for language. The references from the articles were screened and in this way additional articles were obtained. Using the search terms, 382 possible titles were screened. Of these, 334 were not relevant as they discussed sports injuries in general, stress fractures, compartment syndromes or other topics. The 48 relevant titles were screened for related titles in the references. In total, 110 references were found, of which 104 articles could be obtained. Articles were judged using the Institute for Quality of Healthcare (CBO [Centraal Begeleidings Orgaan]) classification system[15] (table I) and methodological quality and level of evidence were assessed. Methodological quality status (A1, A2, B, C, D) and level of evidence status (1, 2, 3, 4) were assessed (see tables II and III). The assessment was done independently by two researchers (MM and MS). If methodological quality and level of evidence were scored differently, a third author (AW) made the final decision (on two occasions). Randomized controlled studies on the prevention and treatment of MTSS were also assessed using the Delphi scoring list[39] (table IV and V). This is a list of criteria for quality assessment of randomized clinical trials when conducting systematic reviews. This list contains nine points and each was scored as being present or not. The maximal score for the Delphi list is nine points. 2. Aetiology 2.1 Functional Anatomy
There is much controversy about the anatomical basis for MTSS. Post-mortem studies have been performed to examine the relationship ª 2009 Adis Data Information BV. All rights reserved.
A2: Randomized double-blind clinical comparing study of good quality and size B: Randomized clinical study, with moderate quality and size, or other comparing research (case-control study, cohort study) C: Case series D: Expert opinion Assessment of methodological quality of studies concerning imaging and aetiology A1: Systematic review of at least two independently conducted studies of A2 level
Imaging A2: Research comparing against a gold standard/reference test, with an adequate number of participants B: Research comparing against a gold standard/reference test, with an inadequate number of participants
Aetiology A2: Prospective research with adequate and non-selective follow-up, with control for confounding B: Prospective research with not all criteria mentioned under A2, or retrospective research
Imaging and aetiology C: Case series D: Expert opinion Level of evidence 1: One systematic review (A1) or at least two independently conducted studies of A2 level (strong evidence) 2: One study of A2 level, or at least two independently conducted studies of B level (moderate evidence) 3: One study of B or C level (limited evidence) 4: Expert opinion (no evidence)
between the location of the pain and the anatomical structures. In these studies the distal attachments of different leg muscles were compared with the site of symptoms in MTSS. Michael and Holder[49] dissected 14 specimens and found fibres of the soleus muscle but not the posterior tibialis muscle on the posteromedial tibial border. Saxena et al.[50] dissected ten cadavers and found that the distal attachment of the tibialis posterior muscle was 7.5 cm proximal to the medial malleolus. He concluded from this that the tibialis posterior muscle caused MTSS. Sports Med 2009; 39 (7)
526
ª 2009 Adis Data Information BV. All rights reserved.
Table II. Study characteristics and quality scores of studies involving imaging Study design
Inclusion criteria
Imaging type
No. of subjects
Population/type of activity
Outcome
Methodological quality
Level of evidence
Holder and Michael[16] (1984)
Prospective cohort
Pain on palpation of middle and distal posteromedial tibial border
Bone scan
10
Athletes, 50% M, 50% F; 6 running, 2 hockey, 1 ballet, 1 basketball, 16–31 y
9 scans abnormal uptake, 1 normal
B
2
Chisin et al.[17] (1987)
Prospective cohort
Not clearly stated
Bone scan
171 scanned with suspicion of stress fracture
Male soldiers, 18–21 y
171 bone scans: 53% sharply defined abnormality, stress fracture, 35% irregular poorly defined uptake, 12% normal
B
2
Batt et al.[18] (1998)
Prospective cohort
Exercise-induced lower leg pain, pain on palpation >5 cm on posteromedial tibial border
MRI/bone scan/x-ray
23: 41 symptomatic tibias, 4 asymptomatic athletes
Athletes and students, 14–58 y; 48% F, 52% M
x-Ray: 9% periosteal elevation; bone scan: 88% tibias abnormal; MRI: 83% abnormal
B
2
Gaeta et al.[19] (2005)
Case control
Lower leg pain <1 month; x-ray normal, clinical exam not stated
MRI/bone scan/CT scan
42: 50 tibias; 10 asymptomatic
Recreational and competitive athletes, 16–37 y; 38% F, 62% M
MRI: 88% abnormal; CT: 42% abnormal; bone scan: 74% abnormal; MRI and CT normal in asymptomatic athletes
B
2
Gaeta et al.[20] (2006)
Case control
Exercise-related pain at posteromedial tibial border
Highresolution CT scan
20 asymptomatic athletes, 10 asymptomatic non-athletes, 11 symptomatic (14 tibias)
Distance runners, 18–26 y; 32% F, 68% M
Asymptomatic non-athletes: 95% tibias normal. Asymptomatic athletes: 45% abnormal; all 14 painful tibias: abnormal
B
2
Fredericson et al.[21] (1995)
Retrospective cohort
Runners with tibial pain with confirmation of MTSS; tibial stress reaction of tibial stress fracture on bone scan
MRI/bone scan
14: 18 tibias
Runners (track, hurdles, distance runners), 18–21 y; 21% M, 79% F
Grade I and II: periosteal oedema and bone marrow oedema on T2 weighted; grade III and IV: periosteal oedema on T2, marrow oedema on T1 and T2. Correlation bone scan/MRI in 78%
B
3
Continued next page
Moen et al.
Sports Med 2009; 39 (7)
Study (year)
Study design
Inclusion criteria
Imaging type
No. of subjects
Population/type of activity
Outcome
Methodological quality
Level of evidence
Arendt et al.[22] (2003)
Retrospective cohort
Athletes who underwent MRI with suspicion of stress fracture
MRI
26
Athletes; basketball, running, football, gymnastics, ice hockey, track, tennis, softball; 31% M, 69% F
The more severe the lesion on MRI the longer the time to return to sport
B
3
Rupani et al.[23] (1985)
Case series
Not clearly stated
Bone scan
44
Recreational and competitive athletes, 11–72 y; F/M ratio not clearly stated
Distinguishing tibial stress fractures and MTSS is possible with bone scan
C
3
Nielsen et al.[24] (1991)
Case series
Pain along the posteromedial border
Bone scan/x-ray
22: 29 tibias
Male soldiers (age unknown)
x-Ray: 45% abnormal; bone scan: 83% abnormal uptake; 17% normal
C
3
Anderson et al.[25] (1997)
Case series
Activity-related lower leg pain and tenderness on palpation along the posteromedial tibia
MRI/x-ray
19
Competitive and recreational athletes, 17–54 y; 58% F, 42% M
37% MRI normal, 26% MRI periosteal fluid; 26% MRI bone marrow oedema, 11% stress fracture; x-ray: 5/5 normal
C
3
Matilla et al.[26] (1999)
Case series
Medial tibial pain within 500 m of marching; x-ray normal, pain >5 cm along tibial shaft
MRI
12: 14 tibias
Male soldiers, 17–25 y
93% periosteal oedema; 29% intraosseous bright signal and periosteal oedema
C
3
Aoki et al.[27] (2004)
Case series
Pain in the middle or distal portion of the medial side of the leg; normal xray
MRI
14 MTSS, 8 stress tibial fracture
Athletes (runners, basketball, volleyball, kendo, soccer players), 13–33 y; 59% M, 41% F
14/14: linear abnormally high signal along posteromedial border, 50% abnormally high signal of bone marrow, 36% both abnormal signals seen. After 4 wk, with continued exercise, MRI signals diminished in 5 patients
C
3
CT = computed tomography; F = female; M = male; MRI = magnetic resonance imaging; MTSS = medial tibial stress syndrome; T1 = T1 weighted; T2 = T2 weighted.
527
Sports Med 2009; 39 (7)
Study (year)
Medial Tibial Stress Syndrome
ª 2009 Adis Data Information BV. All rights reserved.
Table II. Contd
528
ª 2009 Adis Data Information BV. All rights reserved.
Table III. Study characteristics and quality of studies concerning intrinsic risk factors Study (year)
Inclusion criteria
No. of subjects
Population
Risk factors (specification of determinant)
Outcome
Methodological Level of quality evidence
DeLacerda[28] Prospective (1980) cohort
Pain along the posteromedial aspect of the tibia
81
Female, physical education students, 18–21 y
Navicular displacement weight bearing/ non-weight bearing
Incidence MTSS 37%. Navicular drop 8.90 – 2.89 mm in MTSS group, control 5.56 – 2.32 mm
A2
2
Bennett Prospective et al.[3] (2001) cohort
Pain with palpation 125 over the distal 2/3 of the posterior medial tibia
Cross-country runners, 14–17 y; 46% M, 54% F
Navicular drop test
Navicular drop test (p = 0.01), female sex (p = 0.003)
A2
2
Burne et al.[29] Prospective (2004) cohort
At least 1 wk medial 158 tibial pain on exertion and >10 cm pain on palpation at distal 2/3 of posteromedial tibia
Military cadets, 17–21 y; 77% M, 23% F
Men only: greater internal and external hip ROM, leaner calf girth
Incidence MTSS 15%. A2 Incidence 15% F, 10% M. Greater internal and external ROM (p < 0.05), leaner calf girth (p = 0.04)
2
Prospective cohort
Pain, due to 125 exercise along the posteromedial tibial border, on palpation diffuse >5 cm
Naval recruits, 17–35 y; 75% M, 25% F
Female sex (RR 2.03), Incidence MTSS 36%. A2 more pronated foot Incidence 53% F, 28% M type (RR 1.70)
2
Plisky et al.[30] Prospective (2007) cohort
Pain along the distal 105 2/3 of the tibia exacerbated with repetitive weightbearing activity
Cross-country runners, 14–19 y; 56% M, 44% F
Higher BMI (RR 5.0)
Incidence MTSS 15%. 4.3/1000 athletic exposures (F), 1.7/1000 athletic exposures (M)
A2
2
Yates and White[4] (2004)
Study design
Prospective cohort
Exercise-related pain along the posteromedial side of the tibia for at least 5 cm with diffuse pain on palpation
146
Collegiate athletes from NCAA division I and II, 20 – 1.7 y; 45% M, 55% F
Athletic activity <5 y, previous history of MTSS/ stress fracture, use of orthotics
Incidence MTSS 20%. A2 Incidence 11% F, 31% M
2
Gehlsen and Seger[32] (1980)
Case control
Not clearly stated
10 symptomatic, 10 control
Female athletes, age not stated; 10 symptomatic, 10 control
Increased plantar flexor strength. Decreased inversion flexibility (right ankle), increased angular displacement (Achilles tendon/calcaneus)
Increased plantar flexor strength (p < 0.05). Decreased ankle inversion (p < 0.05). Increased angular displacement (p < 0.05)
3
B
Continued next page
Moen et al.
Sports Med 2009; 39 (7)
Hubbard et al.[31] (2009)
Study design
Inclusion criteria
No. of subjects
Population
Risk factors (specification of determinant)
Outcome
Methodological Level of quality evidence
Viitasalo and Kvist[33] (1983)
Case control
Regular or long lasting pain on the medial border of the distal 2/3 of the tibia
13 controls, 13 with frequent and long-lasting MTSS, 22 slight MTSS
Male distance runners, judo, soccer, skiing, boxing, basketball. Age: control 30.6 – 7 y, frequent 23.8 – 7 y, slight 19.8 – 5 y
Increased mobility of inversion, eversion and sum. Achilles tendon angle displacement smaller during full support phase
Passive mobility inversion (p < 0.01), eversion (p < 0.05), sum (p < 0.001). Angular displacement during full support (p < 0.01)
B
3
Sommer and Vallentyne[34] (1994)
Case control
Regular or long lasting pain on the medial border of the distal 2/3 of the tibia
25 subjects: 15 controls, 10 cases of which 4 bilateral
Amateur folk dancers, 15–25 y; 80% F, 20% M. 10 previously diagnosed with MTSS, 15 controls
Combination of forefoot and hindfoot varus alignment. Standing foot angle <140
Forefoot and hindfoot varus (p = 0.047). Standing foot angle <140 (p = 0.0001)
B
3
Madeley et al.[35] (2007)
Case control
Exercise-induced leg pain of the posteromedial border of the tibia. Pain on palpation >40 mm on 100 mm visual analogue pain scale
30 symptomatic (with 59 painful tibias), 30 controls
Athletes, 17–47 y; MTSS group 53% M, 47% F; control group 53% M, 47% F
Standing heel-rise repetitions
Standing heel rise (p < 0.001)
B
3
Tweed et al.[36] (2008)
Case control
12 control, Exercise-induced 28 with MTSS pain on the posteromedial border of the tibia for at least 4 cm and pain on palpation
Early heel lift, Runners, 18–56 y; abductory twist during MTSS group: 43% F, 57% M; control group: gait, apropulsive gait 42% F, 58% M
Early heel lift (EOR 27), abductory twist (EOR 123), apropulsive gait (EOR 823)
B
3
Bandholm et al.[37] (2008)
Case control
Exercise-induced pain on the posteromedial tibial border and pain on palpation >5 cm
15 control, 15 with MTSS
Athletes, 20–32 y; MTSS group 60% F, 40% M; control 60% F, 40% M
Larger navicular drop and MLAD during stance. Larger MLAD during gait
Larger navicular drop during stance (p = 0.046), larger MLAD during stance (p = 0.037), MLAD during gait (p = 0.015)
Taunton et al.[38] (2002)
Retrospective Not clearly stated cohort
2002 running injuries
Runners, mean age MTSS subgroup 30.7 y; 43% M, 57% F
Below average activity history (OR 3.5 M, OR 2.5 F)
Incidence 5% MTSS
B
3
BMI = body mass index; EOR = estimated OR; F = female; M = male; MLAD = medial longitudinal arch deformation; MTSS = medial tibial stress syndrome; NCAA = National Collegiate Athletic Association; OR = odds ratio; ROM = range of motion; RR = relative risk.
529
Sports Med 2009; 39 (7)
Study (year)
Medial Tibial Stress Syndrome
ª 2009 Adis Data Information BV. All rights reserved.
Table III. Contd
Study (year)
No. of subjects
Population
Andrish et al.[2] (1974)
2777
First-year male midshipmen; age not stated
Five groups: ice application; aspirin (acetylsalicylic acid) and ice; phenylbutazone and ice; heel-cord stretching and ice; walking cast
Nissen et al.[40] (1994)
23 experimental, 26 control
Soldiers; age not stated
Johnston et al.[41] (2006)
7 experimental, 6 controls
Soldiers; 18–37 y; sex not stated
a
Intervention
Outcome
Delphi itemsa
530
ª 2009 Adis Data Information BV. All rights reserved.
Table IV. Methodological quality of randomized controlled trials according to the Delphi criteria (treatment) [reproduced from Verhagen et al.,[39] with permission] Total score
Methodological quality
Level of evidence
-
4/9
A2
2
-
-
5/9
A2
+
+
5/9
A2
1a
1b
2
3
4
5
6
7
8
Incidence MTSS 4%. No significant differences
+
+
+
+
-
-
-
-
Two groups: active laser treatment; placebo laser
No significant differences between groups in VAS score and days to return to active duty
+
+
+
-
-
+
+
Two groups: leg orthosis and walkto-run programme; walk-to-run programme
No significant differences between groups in days to recovery (p = 0.575)
+
+
+
-
-
-
-
2
Delphi items (+ indicates ‘yes’, - indicates ‘no’):[39]
1a: Was a method of randomization performed? 1b: Was the treatment allocation concealed? 2: Were the groups similar at baseline regarding the most important prognostic indicators? 3: Were the eligibility criteria specified? 5: Was the care provider blinded? 6: Was the patient blinded? 7: Were point estimates and measures of variability presented for the primary outcome measures? 8: Did the analysis include an intent-to-treat analysis? MTSS = medial tibial stress syndrome; VAS = visual analogue scale.
Moen et al.
Sports Med 2009; 39 (7)
4: Was the outcome assessor blinded?
Population
Andrish et al.[2] (1974)
2777 first-year male midshipmen; age not stated
Five groups: control; heel pad of foam rubber; heel cord stretches; heel pad and stretches; graduated running programme prior to training
Control group 3.0%, heel pad group 4.4%. No significant difference was found
+
+
+ + - - - - - 4/9
A2
2
Bensel and Kish[42] (1983)
2841 army basic trainees; age 16–41 y; 73% M, 27% F
Two groups: 1 hot weather boots; 2 black leather combat boots
1. M 0.27%, F 1.18%; 2. M 0.22%, + F 1.17%. Not significant
+
+ + - - - + - 5/9
A2
2
Bensel[43] (1986)
555 female soldiers; age unknown
Three groups: urethane foam insole; moulded network of lever-like projections attached to material in grid form; standard plastic mesh with nylon
MTSS with different insoles, varying from 5.9% to 7.4%. Not significant
+
+
+ - - - - + - 4/9
A2
2
Schwellnus et al.[44] (1990)
1388 military recruits; 17–25 y; sex not stated
Two groups: neoprene impregnated with nitrogen MTSS: control 6.8%, experimental + bubbles covered with nylon; no intervention 2.8% (p < 0.05)
+
+ - - - - - - 3/9
A2
2
Schwellnus and Jordaan[45] (1992)
1398 male military recruits; age < 25 y
Two groups: 800 mg/day calcium supplementation; no supplementation
MTSS: control 20.4%, calcium group 33.3%. Not significant
+
+
+ - - - - - - 3/9
A2
2
Pope et al.[46] (2000)
1538 male army recruits; age 17–35 y
Two groups: stretching gastrocnemius, soleus, hamstring, quadriceps, hip adductor and hip abductor muscle groups; no stretching. Both groups same physical protocol
MTSS 1.6%. No effect of + stretching on injury risk. LR = 1.24, HR 1.23
+
- + - - - + + 5/9
A2
2
Larsen et al.[47] (2002)
146 military conscripts; Two groups: custom-made biomechanic shoe 145 men and 1 woman; orthose; no intervention age 18–24 y
+
+
+ - + + - + + 7/9
A2
2
Brusho¨y et al.[48] (2008)
1020 military recruits training for Royal Danish Life Guard; 19–26 y; sex not stated
Two groups: strength, coordination and stretching MTSS: leg training 4.5%, upper + exercises of the legs; strength and stretching body training 4.9%. Not significant exercises of upper body
+
+ - + - + + - 6/9
A2
2
a
Intervention
Outcome (incidence)
Delphi itemsa
Study (year)
1a 1b 2 3 4 5 6 7 8
MTSS: control group 38%, intervention group 9% (p = 0.005). RR 0.2; cost per prevented case $US101 (2002 values)
Total Methodological score quality
Level of evidence
Medial Tibial Stress Syndrome
ª 2009 Adis Data Information BV. All rights reserved.
Table V. Methodological quality of randomized controlled trials according to the Delphi criteria (prevention) [reproduced from Verhagen et al.,[39] with permission]
Delphi items (+ indicates ‘yes’, - indicates ‘no’):[39]
1a: Was a method of randomization performed? 1b: Was the treatment allocation concealed? 2: Were the groups similar at baseline regarding the most important prognostic indicators? 3: Were the eligibility criteria specified? 5: Was the care provider blinded? 6: Was the patient blinded? 7: Were point estimates and measures of variability presented for the primary outcome measures? 8: Did the analysis include an intent-to-treat analysis? F = female; HR = hazard ratio; LR = likelihood ratio; M = male; MTSS = medial tibial stress syndrome; RR = relative risk.
531
Sports Med 2009; 39 (7)
4: Was the outcome assessor blinded?
Moen et al.
532
Beck et al.[51] dissected 50 legs and concluded that if a traction was implicated in MTSS, the soleus muscle and the flexor digitorum longus muscle rather than the tibialis posterior muscle could be involved. During dissection, no fibres of the tibialis posterior muscle were found on the distal half of the posteromedial border of the tibia. In the upper half of the distal tibia, fibres of the soleus muscle and flexor digitorum longus muscle were abundant on the medial border. While MTSS complaints are commonly felt in the distal third of the tibia, few muscle fibres of the soleus muscle or any other muscle were found at this site.[51] Garth and Miller[52] concluded, after performing a case-control study in 17 athletes, that the flexor digitorum longus muscle caused the complaints. In the symptomatic group he found a decreased flexion range of motion of the second metatarsophalyngeal joint and weakness of the toe flexors. He suggested that this was caused by permanent increased activity of the flexor digitorum longus muscle. Traction of the above-stated muscles on the periosteum is thought, by some authors, to cause MTSS. The traction explanation was first published in the 1950s.[7] This states that complaints are due to repeated traction on the periosteum of the fibres of the tibialis posterior, soleus or flexor digitorum longus muscles. However, symptoms are not always felt at the site of distal attachment of the tibialis posterior, soleus and flexor hallucis longus muscles.[51] The traction explanation has only recently been investigated. Using three cadaver specimens, traction on the periosteum during soleus, posterior tibial and flexor digitorum longus activity was measured.[53] As tension on the tendons of the aforementioned muscles was increased, strain in the tibial fascia, which in our opinion refers to the periosteum, increased in a linear manner. 2.2 Biomechanics
Several explanations for the development of MTSS are found in the literature, of which the traction explanation is one. Another explanation for which much evidence exists states that repeated tibial bending or bowing causes MTSS.[54,55] ª 2009 Adis Data Information BV. All rights reserved.
This mechanism has similarities to the aetiology of a tibial stress fracture.[56] Animal studies showed that repeated bending causes adaptation of the tibia, predominantly at the site where bending forces are the greatest.[57,58] The site of most profound bending is where the tibial diaphysis is narrowest[59] – approximately at the junction of the middle and distal thirds. The goal of adaptation is to strengthen the bone to resist future loading. The adaptation is described in Wolff’s law and the Utah paradigm;[60-64] loads on bones cause bone strains that generate signals that some cells can detect and to which they or other cells can respond. Normally, bones can detect and repair small microdamage caused by strains that stay below the microdamage threshold. Strains above the threshold can cause enough microdamage to escape repair and accumulate.[62] The first clinical study to provide evidence for an altered bending mechanism in MTSS was provided by Franklyn et al.[65] in a recent cohort study. Tibial scout radiographs and crosssectional computed tomography (CT) were used to study bone characteristics of aerobic controls, MTSS subjects and subjects with tibial stress fractures. These authors showed that male subjects with MTSS and tibial stress fractures had a smaller cortical area than aerobic controls. They also calculated that aerobic controls were better adapted to axial loading, torsion, maximum and minimum bending rigidity, and pure bending than subjects with MTSS and tibial stress fractures.[65] In addition, animal and human studies showed that diminished muscle forces negatively influence the bone adaptation process, when weaker muscles opposing tibial bending allow more bending to occur.[66-69] A recent in vivo study showed greater tibial strain when muscles were fatigued.[70] A combination of the traction and bony overload explanation is another hypothesis. The adaptation to loading of the tibia is further challenged by the traction of the soleus and flexor hallucis longus muscles on the periosteum.[56] 3. Histology Histological evidence for periostitis is sparse. Two studies from the 1980s describe three Sports Med 2009; 39 (7)
Medial Tibial Stress Syndrome
patients with inflammation or vasculitis found in the fascia after biopsy.[10,49] In the study by Michael and Holder[49] a thickened periosteum was seen, and termed ‘periostitis’. The study by Mubarak et al.[10] showed two patients with microscopic inflammation and vasculitis of the periosteum. In larger studies, inflammatory cells were not often found in the periosteum.[71,72] Inflammatory changes were found in the crural fascia in 13 of 33 athletes upon biopsy.[72] In the same specimens, one biopsy sample showed evidence of plasma cell infiltration surrounding wide lymphatics in the periosteum, along with a thickened periosteum and increased osteoblast activity. This was also found by Bhatt et al.,[71] who also found fewer osteocytes compared with normal bone, although this finding just failed to achieve statistical significance.[71] They did not describe the activities of their patient population. Evidence in recent literature is accruing that osteocytes play a major role in mechanotransduction, a mechanism through which bone senses mechanical stimuli.[73,74] Osteocytes probably promote bone remodelling in response to a direct mechanical stimulus or to bone microdamage.[75] In bone remodelling, apoptosis of osteocytes is seen and this apoptosis may influence osteoclast formation and/or function.[76] In patients with MTSS, low regional tibial bone density has been found compared with healthy athletes.[77] Bone density in the mid- to distal tibia, measured by dual energy x-ray absorptiometry (DXA), was 23% – 8% (mean – SD) less in patients with MTSS. The bone density regained normal values when the athletes had recovered after a mean of 5.7 years (range 4–8 years).[78] 4. Patient Examination 4.1 History
Patients with MTSS present with exerciseinduced leg pain. The pain is located along the posteromedial border of the tibia, usually in the middle or distal thirds. Initially symptoms are present on starting activity and subside with continued exercise, but later on pain continues to ª 2009 Adis Data Information BV. All rights reserved.
533
be present during activity. If symptoms worsen, then the pain can be felt even after the activity ceases.[79,80] This has also been described in stress fractures, so the physician should be cautious when interpreting this symptom. In severe cases, even performing activities of daily living will provoke symptoms. 4.2 Physical Examination
Few articles were found concerning physical examination and MTSS. V. Ugalde has performed non-published research (personal communication 2006). In this research, an attempt was made to determine the sensitivity and specificity of physical examination tests. Symptomatic athletes and control athletes were included. The gold standard in this study was bone scintigraphy. Three tests were examined: diffuse posteromedial pain on palpation, pain on hopping, and pain on percussion. Diffuse posteromedial pain on palpation was the most sensitive test. During physical examination, pain is present on palpation of the distal two-thirds of the posteromedial tibial border. Mild swelling of the tibia can sometimes be present.[79-81] The risk factors for MTSS are described in section 5 and should be considered during clinical examination. The differential diagnosis of exercise-induced leg pain consists of medial tibial stress syndrome, tibial stress fracture, exertional compartment syndrome, and to a lesser extent popliteal artery entrapment and nerve entrapment.[81] It is our opinion that differentiation between MTSS, tibial stress fracture and exertional compartment syndrome can usually be accomplished without additional imaging. Patients presenting with exertional compartment syndrome complain of cramping, burning or aching pain, and tightness in the leg on exercising. A tight feeling in the muscles and sometimes neurological symptoms such as sensory abnormalities can also be present. Palpation at rest is usually not painful. During exercise the leg is painful, but upon stopping the pain disappears quickly. The diagnosis can be confirmed by intracompartmental pressure measurements.[81] In the 1970s and 1980s some thought that MTSS was caused by elevated leg Sports Med 2009; 39 (7)
Moen et al.
534
compartment pressures. Puranen and Alavaikko[82] studied this in 1981 by measuring the pressure in 22 patients with pain on the medial side of the leg. They found that on exertion, patients had significantly higher increases in pressure than controls. On exertion the pressure ranges in patients and controls were 70–150 mmHg and 15–30 mmHg, respectively. Other researchers failed to find elevated pressure.[10,83,84] In one study, 14 track runners with MTSS showed no elevated pressure present in any compartment.[84] In a series of 12 patients with MTSS, compartment pressures were measured and compared with diagnosed chronic compartment syndromes. In the MTSS group, pressures were lower compared with diagnosed compartment syndrome during exercise (84 mmHg [mean value] vs 112 mmHg, respectively).[10] In 12 patients with MTSS the pressures during exercise were compared with pressures in the compartment syndrome. Values in the compartment syndrome group were higher (28 mmHg and 70 mmHg, respectively).[83] All four studies examining compartmental pressure in MTSS examined relatively few patients, and were of poor methodological quality. Some claim that medial tibial stress syndrome and compartment syndrome may coexist, but apart from the sole study of Puranen and Alavaikko[82] in 1981, no evidence exists. The differentiation between stress fracture and MTSS can sometimes be challenging, especially since radiographs for stress fracture can be false negative with sensitivities as low as 26–56%.[23,85,86] In stress fractures, pain is usually more focal, while in MTSS the pain is more diffuse. Also, night pain and pain on percussion are not usually present in MTSS. Evidence has shown that in persons with stress fractures, plain radiographs are often normal in the first few weeks and may later show callus formation.[6] Bone scintigraphy and magnetic resonance imaging (MRI) are widely used to confirm the diagnosis.[87] 4.3 Imaging
There is a fair amount of literature on MTSS and imaging. In most of the imaging studies the clinical diagnosis is used as the gold standard ª 2009 Adis Data Information BV. All rights reserved.
when establishing sensitivity and specificity of imaging modalities.[18-20] The fact that history and physical examination is used as the gold standard confirms that the diagnosis is made clinically and that the role of additional investigations is limited. Table II describes the study characteristics, methodological quality and results of the imaging studies. 4.3.1 Radiograph
Imaging MTSS with radiograph is not appropriate, with most authors reporting normal radiographs.[25,27,77,88] Callus formation is seldom seen on the medial side of the tibia. In one study, four of 46 patients with pain on palpation for at least 5 cm along the posteromedial tibia showed periosteal elevation on radiograph.[18] In other research describing callus formation, the inclusion criteria for the study were less clear.[10,49] 4.3.2 Bone Scan
In 1984, Holder and Michael[16] were the first to examine MTSS with three-phase bone scans (angiograms, blood-pool images and delayed images) in a prospective study. On delayed images, longitudinal tibial lesions of the posterior cortex, involving one-third of the length of the bone, were seen (figure 1). They suggested that MTSS was a condition in which the periosteum is irritated and osteoblasts are activated. Some years later other researchers studying different athletic populations reached the same conclusion.[17,24] Prospective studies on bone scans by Batt et al.[18] and Gaeta et al.[19] showed a sensitivity of 74–84%. Batt et al. found a 33% specificity (positive likelihood ratio [LR+]/negative likelihood ratio [LR-] 1.25/0.48). The low specificity is explained by the high number of positive scans in asymptomatic athletes and controls. In 1987 and 1988, respectively, Zwas et al.[89] and Matin[90] developed a grading scale for the severity of abnormalities found on bone scan for bone stress injuries. They divided scintigraphic findings into four or five grades. Although the study of Zwas et al. was aimed at stress fractures of the tibia, the results of his study were later used to distinguish between stress fracture and MTSS. Sports Med 2009; 39 (7)
Medial Tibial Stress Syndrome
535
extremity. None of the regions of abnormal lower extremity uptake was symptomatic at the time of initial evaluation. They were referred to a sports medicine clinic and remained asymptomatic after 8–14 months of follow-up. 4.3.3 Magnetic Resonance Imaging
In the last decade MRI has increasingly been used for studying MTSS. On MRI, periosteal oedema and bone marrow oedema can be seen (figure 2a and b).[25,26] Only two studies were found that prospectively examined the sensitivity and specificity of MRI in MTSS. Researchers found that MRI had a 79–88% sensitivity and 33–100% specificity and LR+/LR- of 1.18/0.64 for the diagnosis of MTSS.[18,19] The 100% specificity Gaeta et al.[19] described is based on ten asymptomatic athletes with no abnormalities on MRI. Fredericson et al.[21] and Arendt and Griffiths[5] both developed a grading system for MTSS on MRI, in which Arendt’s system was modified from Fredericson’s. In this grading system MTSS and stress fracture are separated and the severity is graded. In stress fractures more bone marrow oedema and sometimes a fracture line is seen compared with MTSS. a
b
D Fig. 1. Bone scintigraphy. Arrows show abnormal longitudinal uptake in lateral view (reproduced from Aoki et al.,[27] with permission).
Suggestions of a continuum between MTSS and stress fracture were already made in 1979.[91] Differentiating between these two entities has proved difficult with bone scans.[17,18] Batt et al.[18] found in their prospective study, including mainly dancers and runners, that four out of five asymptomatic athletes had abnormalities on bone scans. Other studies also showed false positive bone scans.[16,17] A study of 100 athletes presenting with back complaints, where bone scans were performed, examined the incidence of abnormalities in the lower leg:[92] 34% of the athletes had abnormalities in the lower ª 2009 Adis Data Information BV. All rights reserved.
Fig. 2. Coronal magnetic resonance image showing an abnormally high signal on (a) the medial side of the tibia and (b) along the medial border and the medial side of the bone marrow (reproduced from Aoki et al.,[27] with permission).
Sports Med 2009; 39 (7)
Moen et al.
536
a
b
Fig. 3. Axial T2 weighted images in asymptomatic runners showing (a) periosteal oedema (straight arrows) and (b) bone marrow oedema (curved arrow) [reproduced from Bergman et al.,[93] with permission].
Arendt and Griffiths[5] found, in a retrospective study, that MRI can estimate the time to return to sport. To estimate this, they used an MRI grading scale, previously developed by Fredericson et al.[21] Grade I (positive short T1 inversion recovery [STIR] image) returned to sport in about 4 weeks; grade II (positive STIR and positive T2 weighted) returned to sport in about 6 weeks. It is questionable whether grading on bone scan and MRI can be compared. Batt et al.[18] found a positive correlation between the two imaging techniques in 23 athletes where both bone scan and MRI were performed. Fredericson et al.[21] found no correlation when the MRI and bone scan were compared in 14 athletes with MTSS. Research from Japan[27] points out that MRI can distinguish between stress fracture and MTSS soon after the beginning of tibial complaints. No MRI scans of patients with MTSS showed a signal extending throughout the whole bone marrow, which was present in stress fractures. In MTSS a linear abnormally high signal along the posteromedial border of the tibia and the bone marrow was seen. This study also showed that five athletes with MTSS, who were followed up by MRI 4 weeks after initial MRI, and who continued sports activity, did not develop a stress fracture. In chronic cases (defined as complaints for >46 months in a study investigating athletes, mainly runners) MRI scans were normal in seven patients.[25] Despite abnormalities found on MRI in symptomatic patients, Bergman et al.,[93] in a study with 21 distance runners, showed that 43% had a tibial stress reaction while asymptomatic ª 2009 Adis Data Information BV. All rights reserved.
(figure 3a and 3b). These runners ran 80–100 km a week for 8 weeks and continued doing this. None of these runners developed complaints. 4.3.4 High-Resolution Computed Tomography Scan
With high-resolution CT scan, Gaeta et al.[19,20] showed osteopenic changes in the tibial cortex and few resorption cavities (figure 4). A case-control study reported a sensitivity and specificity of 42% and 100%, respectively (LR 0.58).[19] In ten asymptomatic non-athlete controls, one tibia showed mild abnormalities (slightly reduced cortical attenuation). In 20 asymptomatic runners, 18 of the 40 tibias showed abnormalities (ranging from slightly reduced cortical attenuation to cortical osteopenia). All
Fig. 4. Axial CT scan showing cortical osteopenia (black arrows) and small resorption cavitations (white arrows) [reproduced from Gaeta et al.,[20] with permission].
Sports Med 2009; 39 (7)
Medial Tibial Stress Syndrome
symptomatic tibias in patients with MTSS showed cortical osteopenia.[20] 4.3.5 Imaging Summary
The diagnosis of MTSS should be made clinically. In cases where the diagnosis is unclear the physician may perform a bone scan or an MRI, which have approximately the same sensitivity and specificity. Compared with these values the sensitivity of CT scanning is lower, with a higher specificity. 5. Risk Factors 5.1 Risk Factor Studies
A number of prospective case-control and retrospective studies have examined intrinsic risk factors. Extrinsic risk factors have been poorly studied. The methodological quality and results of the risk studies are described in table III. One of these intrinsic risk factors is overpronation.[4,32,33] However, the definition of pronation in different articles varies. Pronatory foot type was shown to be a risk factor in a prospective military study by Yates and White (relative risk [RR] 1.70),[4] using the Foot Posture Index.[94,95] Gehlsen and Seger[32] and Viitasalo and Kvist[33] found increased pronation upon heelstrike to be a risk factor in two athlete casecontrol studies. In the study by Gehlsen and Seger,[32] the angular displacement between the calcaneus and the midline of the leg while running was significantly greater (p < 0.01) in the MTSS group compared with the non-MTSS group. Viitasalo and Kvist described the same finding as Gehlsen and Seger.[32] The angle between the lower leg and calcaneus at heel strike was higher for the symptomatic group (p < 0.01). Equivalents of pronation, measured with the navicular drop test and the standing foot angle, have also been studied. Four prospective studies were published examining the navicular drop test (the difference in distance between the lower border of the navicular and the ground – loaded and unloaded).[3,28,30,31] The navicular drop test is an indicator of midfoot pronation. Attention to the navicular prominence is also paid in the Foot Posture Index.[94,95] The navicular drop test was ª 2009 Adis Data Information BV. All rights reserved.
537
measured in the study by Bennett et al.[3] of 125 runners. The mean drop distance in runners with complaints was 6.8 mm (– 3.7 mm), compared with 3.7 mm (– 3.3 mm) in the asymptomatic group (p = 0.003). In the second study, a significant correlation was found between navicular tuberosity displacement and the incidence of MTSS (8.9 – 2.9 compared with 5.6 – 2.3 mm) [p < 0.01].[28] A recent case-control study conducted among athletes showed a significant difference (p = 0.046) in navicular drop between loaded and unloaded groups (MTSS group 7.7 – 3.1 mm, control group 5.0 – 2.2 mm).[37] A third and fourth prospective study failed to find a significant relationship between navicular drop and MTSS.[30,31] The standing foot angle measures the angle between medial malleolus, navicular prominence and first metatarsal head. Sommer and Vallentyne[34] found that a standing foot angle <140 was predictive of MTSS (p < 0.0001). The 140 cut-off value was chosen because this led to the best sensitivity and specificity (71.3% and 69.5%, respectively). Recently, two case-control studies examined the functional foot posture in MTSS patients during gait[37] and while running.[36] Using threedimensional gait analysis, the group of recreational athletes with MTSS showed increased medial longitudinal arch deformation during gait compared with healthy controls (p = 0.015).[37] This study also showed increased medial longitudinal arch deformation upon standing compared with the controls. In the study by Tweed et al.,[36] athletes were videotaped during running with and without shoes. Of the variables tested during running, three were significantly different between the groups: early heel rise (p = 0.003; estimated odds ratio [OR] 27), abductory twist of the forefoot (p = 0.003; estimated OR 123) and apropulsive gait (p < 0.001; estimated OR 827). The role of passive inversion and eversion was investigated in a case-control study by Viitasaalo and Kvist[33] among male athletes, showing increased passive inversion (19.5 – 8.6) and eversion (10.7 – 4.4) in the ankle to be an intrinsic risk factor (p < 0.05). The inversion and eversion were measured manually and repeatedly. The correlation coefficient for this measurement was 0.84. Sports Med 2009; 39 (7)
Moen et al.
538
Although suggested in the literature, reduced ankle dorsiflexion has not been shown to be an intrinsic risk factor in a prospective study.[29] Ankle dorsiflexion for males and females was 32 and 29, respectively, in the case group and 31 and 27 in the control group (p > 0.05). Another prospective study showed that increased plantar flexion range of motion was associated with MTSS (p = 0.004). This study was conducted among collegiate athletes.[31] A case-control study published in 1980 reported significantly increased plantar flexion strength values (p < 0.05), using cable tension procedures, in ten athletes with MTSS compared with ten healthy athletes.[32] In an Australian military prospective study by Burne et al.,[29] greater internal and external ranges of hip motion was a risk factor (p = 0.01–0.04 for left and right hip). This was measured with the hip and knee flexed to 90, with the hip rotated until a firm end feel. The extra amount of internal and external hip ranges of motion among patients was 8–12. In the same study[29] the lean calf girth (the maximal calf perimeter corrected for skin thickness) was 10–15 mm less among symptomatic cadets compared with asymptomatic cadets. This finding was only significant among males (p < 0.04). Leaner calf girth may also be biomechanically (see section 2.2) associated with MTSS due to reduced shock-absorbing capacity.[66-69] However, lean calf girth is not strictly correlated with calf muscle strength.[96] In a case-control study, Madeley et al.[35] found a significant difference in the number of heel raises that could be performed. MTSS patients succeeded with 23 repetitions per minute compared with 33 in the controls (p < 0.001). The study demonstrated muscular endurance deficits in athletes with MTSS. A higher body mass index (BMI > 20.2) was shown to be an intrinsic risk factor in the prospective study by Plisky et al. (OR 5.3).[30] The study investigated risk factors in a group of crosscountry runners. Female sex is also an intrinsic risk factor.[3,4,29] In a prospective study of naval recruits in Australia the incidence was 52.9% in females compared with 28.2% in males (RR 2.03).[4] The ª 2009 Adis Data Information BV. All rights reserved.
incidence of MTSS in a group of high school cross-country runners in another prospective study was 19.1% in females and 3.5% in males (p < 0.003).[3] A prospective study among the Australian Defence Force Academy also showed female sex to be a risk factor (MTSS incidence: females 30.6%, males 9.8%; OR 3.1).[29] A retrospective Canadian study found that a below-average activity history (<8.5 years) was an extrinsic risk factor (OR 3.5 in males, 2.5 in females).[38] Prior to analysis of the data, the activity history was divided between >8.5 or <8.5 years. The study evaluated the medical records of 2002 running-related injuries between 1998 and 2000. This study was confirmed by a prospective study that showed that athletes with MTSS had been running less years (5.3 – 1.8 years) than the control group (8.8 – 4.0 years), who did not develop MTSS (p = 0.002).[31] The same prospective study found that athletes with a previous history of MTSS were more likely to develop MTSS than those who had not developed MTSS in the past (p = 0.0001).[31] Risk factors such as increased running intensity, running distance, change in terrain, change of shoes and running with old shoes are often mentioned,[79] but there are no scientific studies supporting these claims. 5.2 Risk Factor Summary
For the intrinsic risk factors, there is level I evidence for excessive pronation and female sex. Level II evidence is available for the risk factors increased internal and external hip ranges of motion, higher BMI, previous history of MTSS and leaner calf girth. 6. Therapy 6.1 Conservative
Only three randomized controlled trials have been conducted on treatment of medial tibial stress syndrome (table IV). All three studies were conducted among military populations. In the first study, by Andrish et al.[2] in 1974, 97 marine recruits who developed MTSS were randomized into five groups. The range of the Sports Med 2009; 39 (7)
Medial Tibial Stress Syndrome
duration of pain prior to inclusion was 1–14 days. Marines in group one did not run until they were pain free, and applied ice over the painful area three times a day. In group two, aspirin (acetylsalicylic acid) [650 mg four times daily] was added for 1 week. In group three, phenylbutazone (100 mg four times daily) was added for 1 week. In group four, additional calf muscle-stretching three times a day for 3 minutes was added. In group five, a plaster walking cast was applied for 1 week. The number of days that the marines were not capable of performing at full activity was recorded. The marines were considered recovered if no pain or tenderness remained or when 500 m running was completed comfortably. The time to recovery for the separate groups was: group 1 – 6.4 days; group 2 – 9.4 days; group 3 – 7.5 days; group 4 – 8.8 days; and group 5 – 10.8 days. The mean time to recovery was 8.6 days. No significant difference was found between the intervention groups. The second study was published in 1994 and had a double-blinded design.[40] Cadets with pain on the posteromedial side of the tibia and pain on palpation of this area were included. The duration of the complaints was not reported. The authors state that other causes besides MTSS of posteromedial pain in the tibia were excluded, without mentioning the specific exclusion criteria. Seventy-two cadets were assessed for inclusion, of which 23 were not eligible or were excluded during the study. The most common reason for exclusion during follow-up was not showing up for treatment. Cadets were randomized into two groups. The first group (n = 26) were treated with a placebo laser probe, while the other group (n = 23) were treated with a functioning laser probe. Both groups received a maximum of six treatments with the probe on the affected part of the tibia. The laser used was a gallium-aluminium-arsenic laser of 830 nm wavelength and 40 mW intensity per 60 seconds per affected centimetre of the tibia. Visual analogue scale (VAS) scores were recorded before every treatment. After 14 days or a maximum of six treatments a physician decided, based on patient history and physical examination, if the cadet could return to duty or not. In the placebo group 19 of 26 cadets (73%) were able to return to duty, and 18 of 23 ª 2009 Adis Data Information BV. All rights reserved.
539
cadets (78%) in the laser group. There was no significant difference between the two groups for return to duty and VAS scores. Just before statistical analysis began, the double-blinded design of the study was changed to single-blinded, due to an accidental breakage of the blinding code. The third study was published in 2006.[41] In this study a leg orthosis was compared with relative rest. The orthosis was an elastic neoprene sleeve with a padded aluminium bar designed to be centred over the most symptomatic portion of the medial leg. Exclusion criterion was any sign of stress fracture on bone scan. Twenty-five soldiers were included, but half of them did not complete the study. Most of them dropped out of the study because of failure to return for follow-up or because of change in permanent training station. Randomization divided the soldiers into two groups: those with and those without a leg orthosis. Both groups followed an identical rehabilitation programme consisting of activity modification and ice massage. Seven days after enrolment in the study a gradual walk-to-run programme was initiated. VAS scores were recorded before and after running. The endpoint was the time until the soldiers could complete running 800 m without pain. Only 13 soldiers completed the rehabilitation programme. Days to completion of the programme were 13.4 – 4.5 (mean – SD) days in the orthosis group and 17.2 – 16.5 days in the control group. These differences were not significant (p = 0.575). In the literature the following treatment regimens are recommended: calf muscle training, using anti-pronation insoles, massage, maintaining aerobic fitness, electrotherapy[97] and acupuncture.[98] Randomized controlled trials or case series studying these treatment options were not found. 6.2 Surgery
The studies reporting surgery were all of poor methodological quality and none had a controlled design. In all of these studies diagnosis was made clinically, and patients with suspected compartment syndrome were excluded. Surgery is sometimes performed when complaints persist after conservative treatment fails. Sports Med 2009; 39 (7)
Moen et al.
540
Different surgical approaches have been described. Some authors[99,100] performed a fasciotomy along the posteromedial border of the tibia using only local anaesthesia. Others[101] used the same technique, but under general anaesthesia. Abramovitz et al.,[102] Detmer[14] and Yates et al.[103] added removing a strip of the periosteum from along the inner border of the tibia. The effect of the operation is thought to be less traction on the periosteum.[102] Regarding the pain, good to excellent results were found in 69–92% of patients: 69% by Yates et al.[103] and 92% by Detmer.[14] Some of the surgical articles report the rate of return to sport[14,100,102,103] The time to return to sport after the operation is poorly stated. Only Detmer[14] states that patients were able to fully resume their sports 3 months after the operation. The results of success in achieving a chance to return to sports have a broad range: 29–93% of patients returned to preoperative sports level. The study by Abramovitz et al.[102] showed 29% return to preoperative sports activity, Holen et al.[100] reported 31%, Yates et al.[103] reported 41%, while Detmer[14] showed a 93% return to preoperative sports level. As mentioned previously, the results should be interpreted with caution, due to the poor study designs. 7. Prevention 7.1 Prevention Studies
Eight randomized controlled trials were found on the prevention of MTSS (table V), all conducted in military populations. The first study, by Andrish et al.,[2] was part of the study that also studied treatment of MTSS. They divided 2777 soldiers randomly into five groups. Group one served as a control group and performed the normal training regimen. The other four subgroups conducted the same training regimen, but to each a preventative intervention was added: the second group wore a heel pad in their shoes; group three performed heel-cord stretching exercises three times daily for 3 minutes; the fourth group performed the same stretches as group three and wore a heel pad; group five entered a ª 2009 Adis Data Information BV. All rights reserved.
gradual running programme 2 weeks before the start of the training schedule and equalled the rest of the groups after the third week of training – they also performed fitness exercises. No significant difference was found between the different groups in incidence of MTSS. In the control group the incidence was 3.0%, with 4.4% in the heel-pad group, 4.0% in the heel cord stretching exercises group, 3.0% in the heel-pad plus heel cord stretching exercises group, and 6.0% in the group with graduated running programme. The second study[42] examined the effect of two kinds of boots in 2841 soldiers over an 8-week period. Training consisted mostly of physical training, although this was not further specified. One boot was constructed of leather, while the other boot had a nose of cotton and nylon (a boot used in tropical environments). The study was conducted to acquire data regarding the effect of the two types of boots on type and frequency of leg disorders among soldiers. The incidence of MTSS, defined as pain and tenderness of the tibia due to overexertion, was the same in both groups. In another study,[43] 555 female soldiers were randomized to wear one of three kinds of insoles. A urethane foam insole and a custom-made insole were compared with a standard insole. During 9 weeks all female soldiers followed the same training programme. There were no significant differences between the groups. A definition of MTSS was not stated. The fourth randomized controlled trial was published in 1990.[44] 1538 soldiers were included, of whom 237 were randomized into an intervention group. They performed 9 weeks of training. The control group wore standard insoles and the intervention group wore neoprene insoles. After 9 weeks of training, 20.4% of the control group had developed MTSS, although this was not defined, compared with 12.8% in the intervention group. This was a significant difference. The fifth study was conducted in 2002.[47] 146 soldiers were randomized to receive standard insoles or a semi-rigid insole, which was handmade and was adjusted per foot. After 3 months of training a significant difference (p < 0.005) was present. Twenty-four (38%) soldiers with the Sports Med 2009; 39 (7)
Medial Tibial Stress Syndrome
standard insole developed MTSS, compared with four (8%) in the intervention group. MTSS was not defined in this study. Schwellnus and Jordaan[45] examined whether calcium supplementation prevented MTSS, for which no definition was given. Of 1398 soldiers, 247 were randomly selected as an experimental group. Before the study started, dietary assessment took place in a selected number of soldiers in the control and experimental groups not yet taking the calcium. Food supplements and calcium intake were calculated. No dietary differences were found. An additional 500 mg of calcium per day was provided to the experimental group. No significant differences were found in the number of patients with MTSS between the experimental and control groups. The seventh preventive study[46] examined preexercise stretching: 1538 army recruits were randomly allocated to stretch or control groups. The stretching protocol consisted of 20 seconds of static stretching for the different lower leg muscles. The study revealed no significant effect on the occurrence of MTSS, which was not further defined. A Danish study[48] examined whether the incidence of MTSS was lowered by a prevention training programme during 12 weeks of military training. Platoons were randomized between two types of training: the prevention training programme and the placebo training programme. The prevention training programme consisted of leg strength and coordination and stretching exercises of the legs, while the placebo training programme consisted of strengthening and stretching exercises of the upper body. MTSS was defined as pain on the medial border of the tibia during running, with pain on palpation of the medial tibial border, not localized to one spot. No significant differences were noted between the training groups. 7.2 Prevention Summary
A number of interventions were studied in the various articles about prevention of MTSS, but of these only a shock-absorbing inlay showed a reduction in the incidence of MTSS in two ª 2009 Adis Data Information BV. All rights reserved.
541
different military studies. For this, level I evidence is available. 8. Discussion A general weakness when reviewing the literature on MTSS is the confusing terminology and the lack of consensus surrounding the definition, which makes comparison of different studies difficult. No widely used definition of MTSS is available in the current literature. Based on the reviewed literature, the following definition of MTSS is suggested: ‘pain felt along the posteromedial border of the tibia’. The pain is aggravated by weight-bearing activity and subsides gradually on stopping. On examination there is recognizable pain on palpation of the posteromedial border over a length of at least 5 cm. This definition distinguishes MTSS from stress fracture, in which the pain is more focal. It is our opinion that the diagnosis can be established clinically. The high prevalence of abnormal imaging studies in asymptomatic athletes means that these techniques should not be used routinely to establish the diagnosis. Several studies show that normal bone remodelling involves resorption of bone before the rebuilding of new bone structures occurs.[60-64] Imaging of tibiae of asymptomatic runners shows abnormalities mimicking the abnormalities found in MTSS.[93] This is thought to represent normal remodelling. From the literature, it is unclear as to whether tibial stress fracture is a continuum of MTSS. In the 1970s Roub et al.[91] were the first to suggest that increased levels of stress to the tibia could result in a spectrum of bony overload. In this spectrum the endstage was a cortical fracture. In the beginning of this spectrum, when bone resorption outpaces bone replacement, MTSS occurs. In both MTSS and stress fractures the same altered bending is present compared with healthy athletic controls.[65] Although a continuum was suggested in the seventies and in our opinion is attractive, no conclusions can be made. In one study[27] athletes with MTSS kept on running after being diagnosed with MTSS. On follow-up MRI scanning, Sports Med 2009; 39 (7)
542
there was no evidence of stress fracture. This could mean that MTSS does not develop into a stress fracture, but also that some symptomatic tibiae healed before fracturing. Possibly, bone variations between individuals determine if one person develops MTSS and the other develops tibial stress facture. If MTSS and stress fracture are not two ends of a continuum of bone injury, then further research is needed to identify the unique pathophysiology of the two conditions. For example, histological samples of MTSS could be studied for microcrack patterns and compared with stress fracture findings. Recently, O’Brien et al.[104] and Raesi Najafi et al.[105] studied the behaviour of microcracks in loaded bones. This behaviour could be compared in MTSS and stress fractures. Also, more studies using high-resolution CT scanning comparing findings between MTSS and stress fractures could be conducted. Recently, micro-CT images were obtained to assess bone microdamage.[106] Slices 10 mm thick could be made with this CT device. Highly detailed images of microdamage in MTSS and stress fractures could possibly be studied. Many controversies surround MTSS. This syndrome has had at least five different names over the past 50 years. Debate still continues as to the underlying cause of MTSS. For decades periostitis caused by traction of the tibialis posterior, flexor digitorum longus or soleus muscles was commonly cited as the mechanism causing MTSS. However, anatomical studies showed that complaints are regularly felt more distal to the most distal attachment of the tibialis posterior, soleus and flexor digitorum longus muscles. Only one study has investigated the role of traction in MTSS and supplied some scientific data to support traction as a possible contributor in the development of MTSS.[53] Recently bony overload of the medial tibia has been shown to be important as the underlying problem. There are four important findings that support the theory that bony overload forms the primary pathophysiological basis for MTSS: (i) on triple-phase bone scans the last phase is abnormal, showing that the bone and periosteum are involved;[16,18] (ii) on high-resolution CT scan the tibial cortex is found to be osteopenic, as can ª 2009 Adis Data Information BV. All rights reserved.
Moen et al.
be seen in patients as well as in asymptomatic athletes as a sign of bone remodelling;[20] (iii) on MRI images, bone marrow oedema as well as a signal along the periosteum can be seen;[19,27] and (iv) in patients with MTSS, bone mineral density is reduced compared with controls,[77] and when symptoms improve, the bone density returns to normal values.[78] Through prospective studies a number of intrinsic risk factors for MTSS have been established. A pronatory foot type with standing is an intrinsic risk factor.[4] Also, an indicator of midfoot pronation, a positive navicular drop test, is an intrinsic risk factor.[3,28,30] The literature suggests, although opinions vary, that excessive pronation leads to increased internal tibial rotation.[107,108] This could cause higher strains in the tibia and may eventually lead to MTSS. Female sex[3,4,29] is another intrinsic risk factor, in which hypoestrogenism and eating disorders probably play a role. It is well known that hypoestrogenism in menstrual irregularities leads to loss of bone mineral density.[109] Eating disorders, independently of hypoestrogenism, lead to altered modulation of the bone turnover under influence of insulinlike growth factor-1 and leptin hormones.[110] A higher BMI[30] will increase tibial loading and bending, leading to pronounced tibial cortex adaptation[62,64] and increased risk for MTSS. Leaner calf girth is associated with MTSS,[29] because the shock-absorbing capacity of the calf muscles is diminished.[66,67,69] No solid explanation is available as to why greater internal and external hip ranges of motion[29] are intrinsic factors. An external risk factor is the previous history of MTSS, possibly due to individual alterations in individual bone remodelling.[31] Little research has been conducted on the treatment of MTSS. Only three randomized controlled trials, published 30 years apart, were found.[2,40,41] The result of these studies is that no intervention proved more valuable than rest alone. The use of common therapies such as massage, strengthening exercises for the calf muscles and anti-pronatory orthotics has never been investigated. Sometimes surgery is performed if conservative treatment fails. The quality of studies Sports Med 2009; 39 (7)
Medial Tibial Stress Syndrome
studying surgery for MTSS is poor. These studies show that surgery can be useful for pain reduction, but only few athletes will return to their preinjury sports level. Eight studies on prevention of MTSS have been published. Three studies investigated the use of different kinds of insoles. Two studies, using semi-rigid orthotics and a neoprene insole, found a significantly lower incidence of MTSS after this intervention. 9. Conclusions MTSS is a common overuse injury affecting many athletes and military recruits worldwide. The use of the definition of MTSS first used by Yates and White[4] is recommended: ‘‘pain felt along the middle or distal third of the posteromedial border of the tibia’’. The pain is aggravated by weight-bearing activity and subsides gradually on stopping. On examination there is recognizable pain on palpation of the posteromedial border over a length of at least 5 cm. It is most probably primarily due to bony overload of the posteromedial tibial border. There is little evidence to support the commonly cited repeated traction-induced periostitis as the primary underlying aetiological factor. Whether or not MTSS and tibial stress fractures are on a continuum is yet to be established and should be investigated further. MTSS is a clinical diagnosis and the prevalence of abnormal findings in asymptomatic subjects means that results of additional investigations should be interpreted with caution. There is level I evidence showing that pronatory foot type and female sex are intrinsic risk factors. There is level II evidence showing that BMI, greater internal and external ranges of hip motion, and calf girth are also intrinsic risk factors. Level II evidence is present also for previous history of MTSS as an extrinsic risk factor. Only three studies have examined the conservative treatment of MTSS. At present there is no evidence that any treatment is superior to rest alone. There is level I evidence that shockabsorbing insoles may help in the prevention of MTSS. ª 2009 Adis Data Information BV. All rights reserved.
543
Acknowledgements No sources of funding were used to assist in the preparation of this review. The authors have no conflicts of interest that are directly relevant to the content of this review. The authors would like to acknowledge the following persons who made substantial contributions: Belinda Beck, Griffith University, Brisbane, QLD, Australia; Viviane Ugalde, Neuromuscular Center of the Cascades, Bend, OR, USA; Fabio Minutoli, University of Messina, Messina, Italy; Stephen Thacker, Centers for Disease Control and Prevention, Atlanta, GA, USA; Yoshimitsu Aoki, Hokushin Orthopaedic Hospital, Sapporo, Japan; Jack Andrish, Cleveland Clinic, Cleveland, OH, USA.
References 1. Clanton TO, Solcher BW. Chronic leg pain in the athlete. Clin Sports Med 1994 Oct; 13 (4): 743-59 2. Andrish JT, Bergfeld JA, Walheim J. A prospective study on the management of shin splints. J Bone Joint Surg Am 1974 Dec; 56A (8): 1697-700 3. Bennett JE, Reinking MF, Pluemer B, et al. Factors contributing to the development of medial tibial stress syndrome in high school runners. Orthop Sports Phys Ther 2001 Sep; 31 (9): 504-10 4. Yates B, White S. The incidence and risk factors in the development of medial tibial stress syndrome among naval recruits. Am J Sports Med 2004 Apr-May; 32 (3): 772-80 5. Arendt EA, Griffiths H. The use of MR imaging in the assessment and clinical management of stress reactions of bone in high performance athletes. Clin Sports Med 1997 Apr; 16 (2): 291-306 6. Lassus J, Tulikoura I, Konttinen Y, et al. Bone stress injuries of the lower extremity. Acta Orthop Scand 2002 Jun; 73 (3): 359-68 7. Devas MB. Stress fracture of the tibia in athletes or ‘‘shin soreness’’. J Bone Joint Surg Br 1958 May; 40B (2): 227-39 8. Clement DB. Tibial stress syndrome in athletes. J Sports Med 1974 Mar-Apr; 2 (2): 81-5 9. Puranen J. The medial tibial syndrome: exercise ischaemia in the medial fascial compartment of the leg. J Bone Joint Surg Br 1974 Nov; 56-B (4): 712-5 10. Mubarak SJ, Gould RN, Lee YF, et al. The medial tibial stress syndrome: a cause of shin splints. Am J Sports Med 1982 Jul-Aug; 10 (4): 201-5 11. Slocum DB. The shin splints syndrome: medical aspects and differential diagnosis. Am J Surg 1967 Dec; 114 (6): 875-81 12. Devas MB. Shin splints, or stress fractures of the metacarpal bone in horses, and shin soreness, or stress fractures of the tibia, in man. J Bone Joint Surg Br 1967 May; 49 (2): 310-3 13. American Medical Association. Standard nomenclature of athletic injuries presented by subcommittee on classification of sports injuries [abstract]. Chicago (IL): AMA, 1966: 122
Sports Med 2009; 39 (7)
544
14. Detmer DE. Chronic shin splints: classification and management of medial tibial stress syndrome. Sports Med 1986 Nov-Dec; 3 (6): 436-46 15. Institute for Quality and Healthcare. Indeling van methodologische kwaliteit van individuele studies: 2006 Jan [online]. Available from URL: http://www.cbo.nl/product/richtlijnen/ handleiding_ebro/article20060207153532/view [Accessed 2009 May 12] 16. Holder LE, Michael RH. The specific scintigraphic pattern of ‘‘shin splints in the lower leg’’: concise communication. J Nucl Med 1984 Aug; 25 (8): 865-9 17. Chisin R, Milgrom C, Giladi M, et al. Clinical significance of nonfocal findings in suspected tibial stress fractures. Clin Orthop Relat Res 1987 Jul; 220: 200-5 18. Batt ME, Ugalde V, Anderson MW, et al. A prospective controlled study of diagnostic imaging for acute shin splints. Med Sci Sports Exerc 1998 Nov; 30 (11): 1564-71 19. Gaeta M, Minutoli F, Scribano E, et al. CT and MRI imaging findings in athletes with early tibial stress injuries: comparison of bone scintigraphy findings and emphasis on cortical abnormalities. Radiology 2005 May; 235 (2): 553-61 20. Gaeta M, Minutoli F, Vinci S, et al. High resolution CT grading of tibial stress reactions in distance runners. AJR 2006 Sep; 187 (3): 789-93 21. Fredericson M, Gabrielle Bergman A, Hoffman KL, et al. Tibial stress reaction in runners: correlation of clinical symptoms and scintigraphy with a new magnetic resonance imaging grading system. Am J Sports Med 1995 Jul-Aug; 23 (4): 472-81 22. Arendt EA, Agel J, Heikes C, et al. Stress injuries to bone in college athletes: a retrospective review of experience at a single institution. Am J Sports Med 2003 Nov-Dec; 31 (6): 959-68 23. Rupani HD, Holder LE, Espinola DA, et al. Three-phase radionuclide bone imaging in sports medicine. Radiology 1985 Jul; 156 (1): 187-96 24. Nielsen M, Hansen K, Holmer P, et al. Tibial peristeal reaction in soldiers: a scintigraphic study of 29 cases of lower leg pain. Acta Orthop Scand 1991 Dec; 62 (6): 531-4 25. Anderson MW, Ugalde V, Batt M, et al. Shin splints: MR appearance in a preliminary study. Radiology 1997 Jul; 204 (1): 177-80 26. Matilla KT, Komu MES, Dahlstrom S, et al. Medial tibial pain: a dynamic contrast-enhanced MRI study. Magn Reson Imaging 1999 Sep; 17 (7): 947-54 27. Aoki Y, Yasuda K, Tohyama H, et al. Magnetic resonance imaging in stress fractures and shin splints. Clin Orthop Relat Res 2004 Apr; 421: 260-7 28. Delacerda FG. A study of anatomical factors involved in shin splints. J Orthop Sports Phys Ther 1980 Fall; 2 (2): 55-9 29. Burne SG, Khan KM, Boudville PB, et al. Risk factors associated with exertional tibial pain: a twelve months prospective clinical study. Br J Sports Med 2004 Aug; 38 (4): 441-5 30. Plisky MS, Rauh MJ, Heiderscheit B, et al. Medial tibial stress syndrome in high school cross-country runners: incidence and risk factors. J Orthop Sports Phys Ther 2007 Feb; 37 (2): 40-7
ª 2009 Adis Data Information BV. All rights reserved.
Moen et al.
31. Hubbard TJ, Carpenter EM, Cordova ML. Contributing factors to medial tibial stress syndrome; a prospective investigation. Med Sci Sports Exer 2009 Mar; 41 (3): 490-6 32. Gehlsen GM, Seger A. Selected measures of angular displacement, strength and flexibility in subjects with and without shin splints. Res Q Excerc Sport 1980 Oct; 51 (3): 478-85 33. Viitasalo JK, Kvist M. Some biomechanical aspects of the foot and ankle in athletes with and without shin splints. Am J Sports Med 1983 May-Jun; 11 (3): 125-30 34. Sommer HM, Vallentyne SW. Effect of foot posture on the incidence of medial tibial stress syndrome. Med Sci Sports Exerc 1995 Jun; 27 (6): 800-4 35. Madeley LT, Munteanu SE, Bonanno DR. Endurance of the ankle joint plantar flexor muscles in athletes with medial tibial stress syndrome: a case-control study. J Sci Med Sport 2007 Dec; 10 (6): 356-62 36. Tweed JL, Avil SJ, Campbell JA. Biomechanical risk factors in the development of medial tibial stress syndrome in distance runners. J Am Podiatr Med Assoc 2008 NovDec; 98 (6): 436-44 37. Bandholm T, Boysen L, Haugaard S, et al. Foot medial longitudinal arch deformation during quiet standing and gait in subjects with medial tibial stress syndrome. J Foot Ankle Surg 2008 Mar-Apr; 47 (2): 89-95 38. Taunton JE, Ryan MB, Clement DB. A retrospective casecontrol analysis of 2002 running injuries. Br J Sports Med 2002 Apr; 36 (2): 95-101 39. Verhagen AP, de Vet HCW, de Bie RA, et al. The Delphi list: a criteria list for quality assessment of randomised clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol 1998 Dec; 51 (12): 1235-41 40. Nissen LR, Astvad K, Madsen L. Low-energy laser treatment of medial tibial stress syndrome. Ugeskr Laeger 1994 Dec; 156 (49): 7329-31 41. Johnston E, Flynn T, Bean M, et al. A randomised controlled trial of a leg orthosis versus traditional treatment for soldiers with shin splints: a pilot study. Mil Med 2006 Jan; 171 (1): 40-4 42. Bensel CK, Kish RN. Lower extremity disorders among men and women in army basic training and effects of two types of boots. Natick (MA): United States Army Natick Research & Development Laboratories, 1983 43. Bensel CK. Wear test of boot inserts: memorandum for the record. Natick (MA): United States Army Natick Research & Development Laboratories, 1986: 1-8 44. Schwellnus MP, Jordaan G, Noakes TD. Prevention of common overuse injuries by the use of shock absorbing insoles. Am J Sports Med 1990 Nov-Dec; 18 (6): 636-41 45. Schwellnus MP, Jordaan G. Does calcium supplementation prevent bone stress injuries? A clinical trial. Int J Sports Nutr 1992 Jun; 2 (2): 165-74 46. Pope RD, Herbert RP, Kirwan JD, et al. A randomised trial of preexercise stretching for prevention of lower limb injury. Med Sci Sports Exer 2000 Feb; 32 (2): 271-7 47. Larsen K, Weidich F, LeBoeuf-Yde C. Can custom-made biomechanic shoe orthoses prevent problems in the back and lower extremities? a randomised controlled interven-
Sports Med 2009; 39 (7)
Medial Tibial Stress Syndrome
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62. 63. 64.
65.
tion trial of 146 military conscripts. J Manipulative Physiol Ther 2002 Jun; 25 (5): 326-31 Brusho¨y C, Larsen K, Albrecht-Beste E, et al. Prevention of overuse injuries by a concurrent exercise program in subjects exposed to an increase in training load; a randomized controlled trial of 1020 army recruits. Am J Sports Med 2008 Apr; 36 (4): 663-70 Michael RH, Holder LE. The soleus syndrome: a cause of medial tibial stress syndrome (shin splints). Am J Sports Med 1985 Mar-Apr; 13 (2): 87-94 Saxena A, O’Brien T, Bruce D. Anatomic dissection of the tibialis posterior muscle and its correlation to the medial tibial stress syndrome. J Foot Surg 1990 Mar-Apr; 29 (2): 105-8 Beck BR, Osterig LR, Oregon E. Medial tibial stress syndrome: the location of muscles in the leg in relation to symptoms. J Bone Joint Surg Am 1994 Jul; 76 (7): 1057-61 Garth WP, Miller ST. Evaluation of claw toe deformity, weakness of the foot intrinsics, and posteromedial shin pain. Am J Sports Med 1989 Nov-Dec; 17 (6): 821-7 Bouche´ RT, Johnson CH. Medial tibial stress syndrome (tibial fasciitis): a proposed pathomechanical model involving fascial traction. J Am Podiatr Med Assoc 2007 Jan-Feb; 97 (1): 31-6 Hayes WC. Biomechanics of cortical and trabecular bone: implications for assessment of fracture risk. In: Mow VC, Hayes WC, editors. Basic orthopaedic biomechanics. New York: Raven Press, 1991: 93-142 Goodship AE, Lanyon LE, McFie H. Functional adaptation of bone to increased stress. J Bone Joint Surg Am 1979 Jun; 61 (4): 539-46 Beck BR. Tibial stress injuries: an aetiological review for the purposes of guiding management. Sports Med 1998 Oct; 26 (4): 265-79 Judex S, Gross T, Zernicke RF. Strain gradients correlate with sites of exercise-induced bone-forming surfaces in the adult skeleton. J Bone Min Res 1997 Oct; 12 (10): 1737-45 Gross TS, Edwards J, McLeod KJ, et al. Strain gradients correlate with sites of periosteal bone formation. J Bone Min Res 1997 Jun; 12 (6): 982-8 Milgrom C, Giladi M, Simkin A, et al. The area moment of inertia of the tibia: a risk factor for stress fractures. J Biomech 1989; 22 (11-12): 1243-8 Cordey J, Gautier E. Strain gauges used in the mechanical testing of bones: part I, theoretical and technical aspects. Int J Care Inj 1999; 30 Suppl. 1; A7-13 Frost HM. From Wolff’s law to the Utah paradigm: insights about bone physiology and its clinical applications. Anat Rec 2001 Apr; 262 (4): 398-419 Frost HM. A 2003 update of bone physiology and Wolff’s law for clinicians. Angle Orthod 2004 Feb; 74 (1): 3-15 Frost HM. From Wolff’s law to the mechanostat: a new ‘‘face’’ of physiology. J Orthop Sci 1998; 3 (5): 282-6 Forwood MR, Turner CH. The response of rat tibiae to incremental bouts of mechanical loading: a quantum concept for bone formation. Bone 1994 Nov-Dec; 15 (6): 603-9 Franklyn M, Oakes B, Field B, et al. Section modulus is the optimum geometric predictor for stress fractures and medial tibial stress syndrome in both male and
ª 2009 Adis Data Information BV. All rights reserved.
545
66.
67. 68. 69. 70.
71.
72.
73. 74.
75. 76. 77.
78.
79.
80.
81.
82.
83.
84.
female athletes. Am J Sports Med 2008 Jun; 36 (6): 1179-89 Paul IL, Murno MB, Abernethy PJ, et al. Musculo-skeletal shock absorption: relative contribution of bone and soft tissues at various frequencies. J Biomech 1978; 11 (5): 237-9 Radin EL. Role of muscles in protecting athletes from injury. Acta Med Scand Suppl 1986; 711: 143-7 Winter DA. Moments of force and mechanical power in jogging. J Biomech 1983; 16 (1): 91-7 Hill DB. Production and absorption of work by muscle. Science 1960 Mar; 131 (3404): 897-903 Milgrom C, Radeva-Petrova DR, Finestone A. The effect of muscle fatigue on in vivo tibial strains. J Biomech 2007; 40 (4): 845-50 Bhatt R, Lauder I, Finlay DB, et al. Correlation of bone scintigraphy and histological findings in medial tibial syndrome. Br J Sports Med 2000 Feb; 34 (1): 49-53 Johnell O, Rausing A, Wendeberg B, et al. Morphological bone changes in shin splints. Clin Orthop Relat Res 1982 Jul; 167: 180-4 Bonewald LF. Mechanosensation and transduction in osteocytes. BoneKey-Osteovision 2006 Oct; 3 (10): 7-15 Han Y, Cowen SC, Schaffler MB, et al. Mechanotransduction and strain amplification in osteocyte cell processes. Proc Natl Acad Sci USA 2004 Nov 23; 101 (47): 16689-94 Nicolella DP, Moravits DE, Gale AM. Osteocyte lacunae tissue strain in cortical bone. J Biomech 2006; 39 (9): 1735-43 Noble B. Microdamage and apoptosis. Eur J Morphol 2005 Jan-Feb; 42 (1-2): 91-8 Magnusson HI, Westlin NE, Nyqvist F, et al. Abnormally decreased regional bone density in athletes with medial tibial stress syndrome. Am J Sports Med 2001 Nov-Dec; 29 (6): 712-5 Magnusson HI, Ahlborg HG, Karlsson C, et al. Low regional tibial bone density in athletes normalizes after recovery from symptoms. Am J Sports Med 2003 Jul-Aug; 31 (4): 596-600 Kortebein PM, Kaufman KR, Basford JR, et al. Medial tibial stress syndrome. Med Sci Sports Exerc 2000 Mar; 32 Suppl. 3: S27-33 Andrish JT. The shin splint syndrome. In: DeLee JC, Drez D, editors. Orthopaedic sports medicine. 2nd ed. Amsterdam: Elsevier, 2003: chapter 29, 2155-8 Edwards PH, Wright ML, Hartman JF. A practical approach for the differential diagnosis of chronic leg pain in the athlete. Am J Sports Med 2005 Aug; 33(8): 1241-9 Puranen J, Alavaikko A. Intracompartmental pressure increase on exertion in patients with chronic compartment syndrome. J Bone Joint Surg Am 1981 Oct; 63 (8): 1304-9 Wallensten R, Eklund B. Intramuscular pressures in exercise-induced lower leg pain. Int J Sports Med 1984 Feb; 5 (1): 31-5 D’Ambrosia RD, Zelis RF, Chuinard RG, et al. Interstitial pressure measurements in the anterior and posterior compartments in athletes with shin splints. Am J Sports Med 1977 May-Jun; 5 (3): 127-31
Sports Med 2009; 39 (7)
546
85. Greaney RB, Gerber FH, Laughlin RL, et al. Distribution and natural history of stress fractures in U.S. Marine recruits. Radiology 1983 Feb; 146 (2): 339-46 86. Kiuru MJ, Pihlajamaki HK, Hietanen HJ, et al. MR imaging, bone scintigraphy and radiography in bone stress injuries of the pelvis and lower extremity. Acta Radiol 2002; 43: 207-12 87. Boden BP, Osbahr DC, Jimenez C. Low-risk stress fractures. Am J Sports Med 2001 Jan-Feb; 29 (1): 100-11 88. Brukner P. Exercise related lower leg pain: bone. Med Sci Sports Exerc 2000 Mar; 32 Suppl. 3: S15-26 89. Zwas ST, Elkanovitch R, Frank G. Interpretation and classification of bone scintigraphic findings in stress fractures. J Nucl Med 1987 Apr; 28 (4): 452-7 90. Matin P. Basic principles of nuclear medicine techniques for detection and evaluation of trauma and sports medicine injuries. Semin Nucl Med 1988 Apr; 18 (2): 90-112 91. Roub LW, Gumerman LW, Hanley EN, et al. Bone stress: a radionuclide imaging perspective. Radiology 1979 Aug; 132 (2): 431-8 92. Drubach LA, Connoly LP, D’Hemecourt PA, et al. Assessment of the clinical significance of asymptomatic lower extremity uptake abnormality in young athletes. J Nucl Med 2001 Feb; 42 (2): 209-12 93. Bergman AG, Fredericsson M, Ho C, et al. Asymptomatic tibial stress reactions: MRI detection and clinical follow-up in distance runners. AJR 2004 Sep; 183 (3): 635-8 94. Redmond AC, Crosbie J, Ouvrier RA. Development and validation of a novel rating system for scoring standing foot posture: the Foot Posture Index. Clin Biomech 2006 Jan; 21 (1): 89-98 95. Keenan AM, Redmond AC, Horton M, et al. The Foot Posture Index: Rasch analysis of a novel, foot-specific outcome measure. Arch Phys Med Rehabil 2007 Jan; 88 (1): 88-93 96. Bamman MM, Newcomer BR, Larson-Meyer DE, et al. Evaluation of the strength-size relationship in vivo using various muscle size indices. Med Sci Sports Exerc 2000 Jul; 32 (7): 1307-13 97. Morris RH. Medial tibial syndrome: a treatment protocol using electric current. Chiropractic Sports Med 1991; 5 (1): 5-8 98. Schulman RA. Tibial shin splints treated with a single acupuncture session: case report and review of the literature. J Am Med Acupuncture 2002; 13 (1): 7-9 99. Ja¨rvinnen M, Niittymaki S. Results of the surgical treatment of the medial tibial stress syndrome in athletes. Int J Sports Med 1989 Feb; 10 (1): 55-7
ª 2009 Adis Data Information BV. All rights reserved.
Moen et al.
100. Holen KJ, Engebretsen L, Grondvedt T, et al. Surgical treatment of medial tibial stress syndrome (shin splints) by fasciotomy of the superficial posterior compartment of the leg. Scand J Med Sci Sports 1995 Feb; 5 (1): 40-3 101. Wallenstein R. Results of fasciotomy in patients with medial tibial stress syndrome or chronic anteriorcompartment syndrome. J Bone Joint Surg Am 1983 Dec; 65 (9): 1252-5 102. Abramowitz AJ, Schepsis A, McArthur C. The medial tibial stress syndrome: the role of surgery. Orthop Rev 1994 Nov; 23 (11): 875-81 103. Yates B, Allen MJ, Barnes MR. Outcome of surgical treatment of medial tibial stress syndrome. J Bone Joint Surg Am 2003 Oct; 85 (10): 1974-80 104. O’Brien FJ, Hardiman DA, Hazenberg JG, et al. The behaviour of microcracks in compact bone. Eur J Morphol 2005 Feb-Apr; 42 (1-2): 71-9 105. Raesi Najafi A, Arshi AR, Eslami MR, et al. Micromechanics fracture in osteonal cortical bone: a study of the interactions between microcrack propagation, microstructure and the material properties. J Biomech 2007; 40 (12): 2788-95 106. Wang X, Masse DB, Leng H, et al. Detection of trabecular bone microdamage by micro-computed tomography. J Biomech 2007; 40 (15): 3397-403 107. Cole GK, Nigg BM, van den Bogert AJ. Transfer of eversion to internal leg rotation in running [abstract]. J Biomech 1994; 27 (6): 659. Presented at International Society of Biomechanics XIV Congress 1993 108. Hintermann B, Nigg BM. Pronation in runners: implications for injuries. Sports Med 1998 Sep; 26 (3): 169-76 109. DeSouza MJ, Williams NI. Physiological aspects and clinical sequelae of energy deficiency and hypoestrogenism in exercising women. Hum Reprod Update 2004 Sep-Oct; 10 (5): 433-48 110. DeSouza MJ, Williams NI. Beyond hypoestrogenism in amenorrheic athletes: energy deficiency as a contributing factor for bone loss. Curr Sports Med Rep 2005 Feb; 4 (1): 38-44
Correspondence: Dr Maarten H. Moen, University Medical Centre Utrecht, Postbus 85500, 3508 GA Utrecht, the Netherlands. E-mail:
[email protected]
Sports Med 2009; 39 (7)
REVIEW ARTICLE
Sports Med 2009; 39 (7): 547-568 0112-1642/09/0007-0547/$49.95/0
ª 2009 Adis Data Information BV. All rights reserved.
Physical Attributes, Physiological Characteristics, On-Court Performances and Nutritional Strategies of Female and Male Basketball Players Gal Ziv1 and Ronnie Lidor1,2 1 The Zinman College of Physical Education and Sport Sciences, Wingate Institute, Netanya, Israel 2 Faculty of Education, University of Haifa, Haifa, Israel
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Physical Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Female Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Male Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Physiological Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Aerobic Profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 Female Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2 Male Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Female Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Male Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3 Female versus Male Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Anaerobic Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Female Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Male Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Agility and Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.1 Female Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.2 Male Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. On-Court Performances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Female Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Male Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Time-Motion Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Heart Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3 Blood Lactate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Nutritional Strategies and Oxidative Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Hydration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Oxidative Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Conditioning for Basketball . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Hormonal Status and the Overtraining Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Five Limitations Observed from the Physical and Physiological Measurements . . . . . . . . . . . . . . . . . . 7. Practical Advice for Basketball and Strength and Conditioning Coaches . . . . . . . . . . . . . . . . . . . . . . 8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
548 549 549 549 551 551 551 552 553 553 554 554 555 555 555 557 557 557 558 558 559 559 560 561 561 561 562 562 563 564 565 566
Ziv & Lidor
548
Abstract
This article reviews a series of studies (n = 51) examining physical attributes, physiological characteristics, on-court performances and nutritional strategies of female and male elite basketball players. These studies included relevant information on physical and physiological variables, such as height, weight, somatotype, relative size, aerobic profile, strength, anaerobic power, agility and speed. Six main findings emerged from our review: (i) differences in physical attributes exist among playing positions and skill levels (e.g. guards tend to be lighter, shorter . and more mesomorphic than centres); (ii) maximum aerobic capacity (VO2max) values of female and male players are 44.0–54.0 and 50–60 mLO2/kg/min, respectively; (iii) male and female players of higher skill levels tend to have higher vertical jump values; (iv) the more skilled female and male players are faster and more agile than the less skilled players; (v) guards tend to perform more high-intensity movements during game play compared with forwards and centres; and (vi) a water deficit of 2% of bodyweight can lead to reduced physical and mental performance during an actual game. Five limitations associated with the testing protocols used in the studies are outlined, among them the lack of a longitudinal approach, lack of tests performed under physical exertion conditions, and lack of studies using a time-motion analysis. In addition, three practical recommendations for the basketball coach and the strength and conditioning coach are presented. It is concluded that the data emerging from these studies, combined with the knowledge already obtained from the studies on physical and physiological characteristics of elite basketball players, should be applied by basketball and strength and conditioning coaches when planning training programmes for elite basketball players.
The game of basketball has established itself as one of the most popular sports in many countries around the globe. Competitive basketball is played not only in North America, where the game was invented and developed, but also on other continents. During the season, elite basketball players – both female and male – practice on a daily basis, often twice a day, play one or two games per week, and take part in international tournaments such as continental and world championships and the Olympic Games.[1] This heavy schedule of practices and games requires careful short- and long-term planning of the players’ training programmes. A training programme for elite athletes (e.g. basketball players) should be composed of three critical phases: preparation, competition and transition.[2,3] In each critical phase, emphasis is placed on four fundamental preparations: physical, technical, tactical and psychological. Among these preparations, the physical preparation is ª 2009 Adis Data Information BV. All rights reserved.
considered to be the major component in most training theories.[4] One of the primary objectives of the physical preparation is to develop the unique fitness components required to attain achievement in a specific sport (e.g. agility, endurance and strength). Relevant information on the physical and physiological characteristics of elite basketball players should be obtained by those professionals – basketball coaches, strength and conditioning coaches, athletic trainers, physiotherapists and sport physicians – who work regularly with the athletes throughout the different phases of the training programme. This information can be appropriately utilized when planning a daily practice, a weekly agenda, or a more long-term programme. It is assumed that such information will help coaches increase their control over the physical and physiological workloads in which the players are engaged, and in turn improve the quality of training. Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
The purpose of this article is threefold: (i) to review a series of studies (n = 51) examining physical attributes, physiological characteristics, oncourt performances and nutritional strategies of female and male basketball players; (ii) to highlight a number of limitations associated with the testing protocols used in the reviewed studies; and (iii) to provide the basketball coach and the strength and conditioning coach with several practical recommendations. 1. Physical Attributes A number of the studies examined the physical attributes of female[5-11] and male[12-19] basketball players. All the studies described the players’ absolute size such as height and weight, while a few also looked at somatotype[5,9,10,16] and relative size.[7] Table I presents a summary of the anthropometric data for both female and male players across the reviewed studies. 1.1 Female Players
Two studies found that the top teams’ players were taller[5,7] and had a longer arm span[7] compared with the bottom teams’ players who took part in the 1994 Women’s World Basketball Championship. These findings suggest that it is possible that absolute size is related to skill level in female players. Studies on female players also showed that percentage body fat (%BF) was higher in centres compared with guards.[6,8] However, centres had a higher fat-free mass (FFM) compared with both guards and forwards. This finding can be explained by the large differences in absolute weight compared with the smaller differences in %BF. A comparison of %BF among female basketball players and female players in other team sports (e.g. teamhandball and volleyball) indicated that basketball players had more %BF than volleyball players, but less %BF than team-handball players.[10] Three studies examined somatotypes – general categories of body type that are independent of absolute body size – in female basketball players.[5,9,10] One study found that guards were more mesomorphic than centres,[5] while another reª 2009 Adis Data Information BV. All rights reserved.
549
ported that guards were more mesomorphic than both centres and forwards.[9] However, Carter et al.[5] suggested that in elite female players, somatotype differences were not great. These finding were in line with the higher %BF found in centres compared with guards in other studies.[6,8] When comparing somatotypes of basketball players with those of volleyball and team-handball players, it was found that basketball players were more mesomorphic than volleyball players, but less than team-handball players.[10] Having a mesomorphic body type along with lower absolute weight can prove useful to guards, who often need to defend against the quickest players of the opposing team and to rapidly transfer the ball from defence to offence while attacking the quickest defenders of the opposite team. The lighter, shorter mesomorphic physique of guards is suitable to the speed and agility required of them. Although female guards were found to be more mesomorphic than centres, centres still showed a higher FFM. When looking at the players’ physique, it is suggested that physical characteristics be considered as a whole, since looking at only one aspect of the players’ physique can be misleading. 1.2 Male Players
One study found that male players of moderate skill level weighed less and were shorter in stature than top-level players.[16] However, another study of male elite players of different skill levels found no differences in height and weight between these players.[17] As suggested by Viviani,[16] it seems that certain physical attributes can negatively or positively affect performance in the game of basketball. One physical variable that was rather consistently different among male players playing different positions was %BF. Out of five studies that reported %BF for male players,[12-15,17] three found that centres had higher %BF than guards,[13,14,17] one found no significant differences among the players playing different positions,[12] and one suggested that centres had lower %BF compared with guards.[15] However, the latter study only sampled one player in the posiSports Med 2009; 39 (7)
Ziv & Lidor
550
Table I. Anthropometric measurements of male and female basketball players. Data shown as mean (SD) Study
Position
Height (cm)
Mass (kg)
%BF
Guards (n = 64)
171.9 (6.1)
66.1 (6.2)
NA
Forwards (n = 57)
181.3 (5.9)
73.3 (5.1)
FFM (kg)
Females Ackland et al.[7] (1997)
Bale[9] (1991)
Centres (n = 47)
189.8 (6.4)
82.6 (8.2)
Guards (n = 7)
162.2 (4.9)
57.9 (6.4)
17.9 (1.1)
Forwards (n = 6)
172.6 (2.7)
63.9 (5.0)
17.9 (2.3)
52.4 (3.2)
Centres (n = 5)
180.0 (4.1)
71.2 (6.4)
18.3 (2.3)
58.1 (4.7)
47.5 (4.9)
Bayios et al.[10] (2006)
All players (n = 133)
174.7 (7.8)
71.5 (10.1)
24.3 (3.6)
53.6 (6.8)
LaMonte et al.[6] (1999)
Guards (n = 18)
169.55 (3.86)
62.15 (5.03)
14.62 (2.58)
52.99 (3.69)
Forwards (n = 19)
179.56 (3.71)
73.61 (6.55)
17.45 (6.06)
60.51 (3.99)
Centres (n = 9)
188.09 (5.46)
79.99 (7.29)
20.79 (4.14)
63.20 (4.6)
Narazaki et al.[20] (2008)
All players (n = 6)
174.2 (9.0)
66.9 (5.8)
19.8 (4.5)
53.7c
Rodriguez-Alonso et al.[11] (2003)
International (n = 14)
180.9 (8.0)
71.7 (7.6)
NA
National (n = 11)
175.1 (6.5)
71.9 (8.7)
Guards (n = 11)
176.5 (4.3)
67.3 (4.8)
Forwardsa (n = 12)
183.3 (NA)
77.9 (NA)
Centres (n = 6)
188.5 (5.2)
81.1 (7.2)
All players (n = 13)
199.5 (6.2)
95.5 (8.8)
11.4 (1.9)
84.5 (NA)c
Guards (n = 8)
183 (4.0)
76.2 (6.4)
6.1 (3.7)
71.6 (NA)c
Forwards (n = 18)
188 (4.0)
77.4 (5.1)
7.8 (4.1)
71.4 (NA)c
Centres (n = 12)
193 (3.0)
Smith and Thomas[8] (1991)
NA
Males Apostolidis et al.[18] (2004) [19]
Ben Abdelkrim et al.
[12]
Cormery et al.
(2007)
(2008)
Guards (n = 26) Forwards (n = 51)
87.2 (5.3) b
185 (0.01)
82.3 (1.66)
b
95.9 (1.15)
b
b
200 (0.01)
78.1 (NA)c
10.4 (7.8) b b
13.7 (0.51)
b
71.02 (NA)c
13.5 (0.35)
b
82.95 (NA)c
14.1 (0.74)
b
95.35 (NA)c
Centres (n = 22)
207 (0.02)
111 (2.42)
Guards
187.4 (5.8) [n = 185]
82.9 (6.8) [n = 185]
8.4 (3.0) [n = 113]
75.8 (8.6) [n = 113]
Forwards
198.4 (3.8) [n = 153]
95.1 (8.3) [n = 152]
9.7 (3.9) [n = 89]
85.5 (8.1) [n = 89]
Centres
205.5 (6.1) [n = 90]
101.9 (9.7) [n = 90]
11.2 (4.5) [n = 53]
90.4 (6.2) [n = 53]
Narazaki et al.[20] (2008)
All players (n = 6)
192.4 (11.7)
91.9 (17.5)
9.7 (5.9)
83.0c
Ostojic et al.[14] (2006)
Guards (n = 20)
190.7 (6.0)
88.6 (8.1)
9.9 (3.1)
79.8 (NA)c
Forwards (n = 20)
200.2 (3.4)
95.7 (7.1)
10.1 (3.2)
86.0 (NA)c
Centres (n = 20)
207.6 (2.9)
105.1 (11.5)
14.4 (5.6)
90.0 (NA)c
Guards
188.0 (10.3) [n = 15]
83.6 (6.2) [n = 15]
10.6 (2.9) [n = 5]
72.9 (6.2) [n = 5]
Forwards
200.6 (5.0) [n = 15]
96.9 (7.3) [n = 15]
9.0 (3.6) [n = 7]
86.6 (6.9) [n = 7]
Centres
214.0 (5.2) [n = 4]
109.2 (13.8) [n = 4]
7.1 (NA) [n = 1]
100.7 (NA) [n = 1]
Guards (n = 14)
185.7 (6.9)
82.0 (8.8)
11.4 (1.7)
72.7 (NA)c
Forwards (n = 22)
195.8 (4.8)
89.4 (7.1)
11.4 (2.3)
79.2 (NA)c
Centres (n = 22)
203.9 (5.3)
103.9 (12.4)
14.4 (3.7)
88.9 (NA)c
[13]
Latin et al.
Parr et al.
[15]
(1994)
(1978)
Sallet et al.[17] (2005)
Continued next page
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
551
Table I. Contd Study
Position
Height (cm)
Mass (kg)
%BF
Viviani[16] (2005)
All playersd
194.2 (6.5)
94.7 (8.7)
NA
Withers et al.[21] (1977)
All playersd
188.8 (7.2)
82.7 (7.3)
16.6 (2.6)
a
Combined average for power forwards and shooting forwards.
b
Data in brackets are standard errors.
c
Not included in original data (calculated from %BF and mass by authors).
d
Not elite players.
FFM (kg) 69.0 (NA)c
%BF = percentage body fat; FFM = fat-free mass; NA = data not available.
tion of centre, and therefore its findings cannot be generalized. Because of the higher absolute weight of centres, it should be noted that although centres tended to have higher %BF, they also possessed higher FFM. As for the somatotypes of male players, one study found that guards were less mesomorphic than other players.[16] A comparison of anthropometric measurements among players from different team sports can help shed light on the specific attributes of basketball players. For example, Withers et al.[21] described various characteristics of male players in basketball, hockey and soccer. They found that basketball players were both taller and heavier than hockey and soccer players. However, this study reported descriptive statistics only. Based on the studies that examined physical characteristics in female and male basketball players, it is observed that differences in height and weight among players playing different positions (e.g. guards, forwards and centres) were the most apparent and consistent. For a specific example, centres were heavier and taller than guards. The tall and heavy build of centres is useful in their physical low-post battles and the necessity to shoot the ball into a rim positioned 3.05 m (10 ft) above ground. 2. Physiological Characteristics 2.1 Aerobic Profile
Although basketball is not an endurance sport per se, having high values of cardiopulmonary functions is important for the player to maintain a high level of activity during the entire game, in both defence and offence. A number of the studies examined cardiopulmonary function in ª 2009 Adis Data Information BV. All rights reserved.
female[8,11,20,22-24] and male[12,14,15,17,18,20,21,25-30] players.
2.1.1 Female Players
The most interesting finding in female players is. the difference in maximum aerobic capacity (VO2max) as reported in studies conducted prior to the enactment in 1972 of Title IX, which is part of an education amendment in the US that prohibited sex discrimination in educational settings, and in those conducted after its enactment. For example, the study by McArdle et al.[22] (published in 1971, before the enactment of Title IX) . found female VO2max values that were only slightly higher than non-athlete college-age female students . (35.5 vs 33.6 mLO2/kg/min, respectively). VO2max was measured at a walking speed of 5.45 km/h and grade was increased until volitional fatigue. Later studies (published from 1979 onwards, after. the enactment of Title IX) found much higher VO2max values, ranging from 44.0 to 54.0 mLO2/kg/min in female basketball players.[8,11,20,23,24] As for the differences between skill levels, Rodriguez-Alonso et al.[11] reported . higher VO2max values for women participating at an international level compared with those at a national level. . Although the exact protocol used to estimate VO2max was unclear, it was apparently estimated . from a progressive run test. The difference in VO2max before and after the enactment of Title IX can be explained, at least in part, by the subsequent increased interest in women’s sports, which resulted in the elevation of scientific interest in regard to female athletes. This was followed by better and more serious training and conditioning programmes for female athletes. Sports Med 2009; 39 (7)
Ziv & Lidor
552
. Different testing protocols for . VO2max can produce different results. The VO2max testing protocol in the study of Smith and Thomas[8] included running at each player’s selected comfortable speed, with grade increased until volitional fatigue. In contrast, Riezebos et al.[23] used . a discontinuous treadmill protocol to assess VO2max. These different testing protocols make it difficult to compare values among studies. In addition, most studies involved increasing grade, whereas basketball players run on level ground. We believe that a better protocol to measure aerobic capacity in basketball players should involve increasing speed . and maintaining level grade. Unlike VO2max, and quite expectedly, maximum heart rate (HRmax) of female players was not different across studies. However, it was found that during an actual game, guards had higher heart rates compared with forwards and centres.[11] This finding can be explained by the high level of work intensity demonstrated by guards during the game compared with that of forwards and centres. However, HR may be influenced by other factors, such as environmental conditions, nutritional status and fitness levels of players, and therefore the interpretation of these results should be performed cautiously. 2.1.2 Male Players
. While VO2max values in females were found to increase in the late 1970s, male players’ values did not change greatly over the last 40. years. However, great variability was found in VO2max values among the male players, ranging from 45.3 – 5.9 mLO2/kg/min in one study[26] to [28] In the former 65.2 mLO . 2/kg/min in another. study, VO2max was measured in a graded test on an electrically braked cycle ergometer, and in the latter study the Bruce protocol was used [31] (for details of this protocol see Whaley . et al., pp. 99-100). The large variability in VO2max can be explained at least in part by the different . protocols used. Nevertheless, in most studies, VO2max values were in the range 50–60 mLO2/kg/min. Court drills are often used in basketball practices to induce skill adaptation in .players. The effects of these drills on HR and VO2 were examined in a study by Castagna et al.[30] of ª 2009 Adis Data Information BV. All rights reserved.
. 14 basketball players. VO2 was measured with a portable metabolic system. Drills included fullcourt games of 2 versus 2, 3 versus 3, and 5 versus 5, with lengths of 3, 4 and 5 minutes. It was found that the intensity of practice was highest. in the 2 versus 2 drills (%HRmax 92.1 – 5.6, %VO2peak 79.0 – 10.7), and was higher . in the 3 versus 3 drills (%HRmax 88.2 – 8.4; %VO2peak 73.5 – 11.6) . than – 9.2%; VO the 5 versus 5 drills (%HRmax, 84.0 2peak . 69.0 – 10.7). Individual HR-VO2 coefficients of determination (r2) were 0.83–0.97. It was suggested that basketball court drills elicit physiological responses that may enhance aerobic fitness. In addition, HR . can be used by coaches as a valid estimator of VO2 and work intensity during basketball drills. A comparison among players playing different positions indicated that guards had the highest . VO2max,[12,14,15,29] reflecting the actual functional requirements of guards during the game. Only one. study, Sallet et al.[17] failed to find differences in VO2max among players playing different positions. Two studies compared aerobic capacity among basketball players and players in other team sports (e.g. soccer . and volleyball). One study[21] described low VO2max values in basketball players compared with both hockey and soccer players, while another study[27] indicated . similar VO2max values among basketball and volleyball players. These studies reported descriptive statistics only. The ability of basketball players to maintain high aerobic capacity during the entire season is of critical importance to both their basketball coach and their strength and conditioning coach. In one study, Tavino et al.[28] found no differ. ences in VO2max as measured in different phases of the season; however, they did not differentiate between starters and bench players. In another study, Caterisano et al.[25] examined physiological variables of starters and reserves during preand post-season phases, and found that .starters were able to maintain a relatively high VO2max, while reserves failed to do so. These findings can be explained by the fact that starters are typically provided with better aerobic stimulation during game play, whereas reserves need to rely mostly on practice sessions, which are usually lower in Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
intensity. However, in practices after game days it is mostly the reserves who practice, and those practice sessions may be intense enough to provide a positive training stimulus for them as well. In the year 2000, several rules of the game were changed by the Fe´de´ration Internationale de Basketball Amateur (FIBA). Three major changes were: (i) a reduction in the time allowed to score a basket from 30 to 24 seconds; (ii) a reduction in the time allowed to cross the median line from 10 to 8 seconds; and (iii) the division of the game into four quarters (4 · 10 min) instead of two halves (2 · 20 min). Cormery et al.[12] compared a number of physiological variables as measured during the two different periods – before the changes made by FIBA and after the realization of these changes. They reported that these. changes were associated with an increased VO.2max in guards, but no significant changes in VO2max in forwards and centres. Although it was suggested that the rule modifications were associated with physiological changes in the players (especially in guards), causation could not be established from this study, since other . factors could have influenced the increased VO2max in guards, among them better training and conditioning programmes, and increased level of competitiveness in the top-level leagues. The ventilatory threshold (VT), which is thought to be related to the anaerobic threshold, is an important measure of aerobic endurance. High VT allows athletes to maintain higher work intensities for longer durations before fatigue appears. Three studies looking at VT found from 50.4% of varying values,[12,18,26] ranging . . [26] VO2max to 77.6% of VO2max.[18] In addition, VT values were similar among players playing different positions, and did not differ after the FIBA rule changes.[12] The variations in VT can be explained by several factors, among them the fitness level of the players, their playing style, and the time of the season in which the tests were administered. Due to the scarcity of data and the varied results that emerged from these studies, more research is needed in order to profile the VT of elite basketball players. ª 2009 Adis Data Information BV. All rights reserved.
553
2.2 Strength
Eleven studies assessed measures of strength in female[8,9,24,32] and male[13,15,25,32-35] basketball players. Various types of tests were used to measure strength, and therefore it was difficult to compare values across studies. Among the tests used were isokinetic tests,[8,23,35] maximal concentric tests[9,13,15,25,33,34] (e.g. bench press one repetition maximum [RM]), muscle endurance[15,24] and isometric tests.[24,32] Variations in strength among players were large, and some of the findings should be treated cautiously. 2.2.1 Female Players
Two studies that examined strength in female players[9,24] represent the difficulty that exists in making comparisons among studies. Both of these studies assessed left and right hand grip, but one assessed isometric strength[24] and the other dynamic strength.[9] No differences were found among players playing different positions in left and right dynamic hand grip strength. As Vaccaro et al.[24] suggested, female players’ strength was clearly superior to that of non-athlete females. Maintaining balance in strength and force production in the limbs and in the antagonist muscles of the same limb are of interest to strength and conditioning coaches, athletic trainers and physiotherapists, since it is possible that imbalances can lead to injury.[36,37] Smith and Thomas[8] examined female basketball players on an isokinetic dynamometer. They found no differences in peak torque of knee flexion and extension between the dominant and the nondominant leg. However, the ipsilateral torque ratio between flexion and extension was outside the generally accepted values. There were large individual variations in this ratio, suggesting that each player should be assessed individually, and that specific exercise programmes should be developed for each player based on their unique characteristics. The scarcity of data regarding the strength of female basketball players is unfortunate, especially since interest in female basketball has been increasing, and training programmes have reached levels similar to those in male basketball. Sports Med 2009; 39 (7)
Ziv & Lidor
554
More research is needed so that strength and conditioning coaches can optimize training sessions when working with elite basketball players. 2.2.2 Male Players
When comparing concentric strength measurements among players playing different positions and at different skill levels, for example in the bench press, relatively similar values were reported. In one study of college players,[13] a mean of 102.3 kg was found, and in another study[34] a mean of 97.2 kg was indicated over 4 years of testing. The highest value was reported in a study comparing starters versus reserves: a mean of 112.7 kg and 111.3 kg, respectively.[25] However, these values were obtained during the pre-season phase. Testing conducted during the post-season phase produced values of 104.2 kg and 98.0 kg for starters and reserves, respectively. The authors suggested that the decrease in bench press 1 RM was due to the limited time the players devoted to strength training during the season. The only measurements of strength in NBA players were performed by Parr et al.[15] In this study, strength was measured on an isokinetic device at an angular velocity of 60 deg/sec. Guards’ bench press strength mean value was 86.8 kg. As expected, this value is lower than the 1 RM values of guards in other studies. Interestingly, the forwards who took part in this study demonstrated relatively high values of strength (101.3 kg), similar to the 1 RM values of college players. This could suggest that the 1 RM values of NBA players may be higher than those of college players. In a study by Latin et al.,[13] forwards showed higher power clean values compared with guards and higher squat values compared with centres, while there were no differences in absolute maximal strength in bench press among players playing different positions. When these values were divided by bodyweight, guards were shown as having the greatest strength. It is possible that upper-body strength is more important for certain playing positions, such as power forward and centre.[34] Training for maintaining or increasing strength throughout the season in players in these ª 2009 Adis Data Information BV. All rights reserved.
positions may be recommended. When comparing bench press and squat performances between basketball and football players, one study found significantly lower values in basketball compared with football players.[33] This was true for both absolute strength and strength relative to bodyweight. One study of isokinetic knee flexion and extension torque in male players failed to reveal significant differences between the dominant and the non-dominant legs.[35] However, unlike in the female players, the study did find a relatively high flexors-extensors ratio in both the dominant and non-dominant legs at an angular velocity of 180 deg/sec in males (84.6% and 83.1%, respectively). It is suggested that those male players possessed relatively strong hamstring muscles. 2.2.3 Female versus Male Players
One study compared force production between male and female players.[32] Quite expectedly, male players showed higher values in isometric leg extension and trunk flexion and extension. In addition, male players took less time to produce maximal force compared with female players. However, when the absolute strength results were adjusted to bodyweight, differences in leg extension force production disappeared. While force production relative to bodyweight was greater in male players for trunk flexion and extension, the differences were smaller than those observed for absolute force production. Moreover, when the absolute values were divided by FFM rather than bodyweight, the differences were even smaller. The discrepancies that remained could be related to sexual distinctions; however, it is possible that differences in volume and intensity of strength training between the female and male players could be contributing factors as well.[32] Since individual variability was the common thread in all studies of strength, the reported values should be carefully considered. Providing individual training essentials should be taken into account by basketball coaches and strength and conditioning coaches, regardless of the mean or expected strength of players in a specific playing position. Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
2.3 Anaerobic Power
It is generally accepted that possessing anaerobic power is essential in most sports, particularly in sport activities requiring the production of work over short periods of time, such as basketball. Various batteries of tests were used in the reviewed studies to assess anaerobic power, and therefore it was difficult to compare results across the studies. Among the tests included were vertical jump, Margaria stair run, Wingate Anaerobic Test (WanT), and an anaerobic power step test. Six studies examining the anaerobic power profile in female[6,8,9,23,32,38] and 12 in male[13-15,17,18,21,28,32-34,38-40] players are reviewed here. 2.3.1 Female Players
Vertical jump is the most prevalent test used to assess anaerobic power in female and male players, since vertical jumps are among the most prevalent acts performed by basketball players in both defence (e.g. blocking and rebounding) and offence (e.g. shooting and rebounding). Various tests assessing vertical jump were used in the reviewed studies, and therefore a wide range of vertical jumping capabilities can be seen. Vertical jump values as reported in these studies are presented in table II. The studies on vertical jump used various protocols. Mean values ranged from 24.8 cm in one study[32] to 48.2 cm in another.[6] However, most studies found values that were above 40 cm. The low value of 24.8 cm was recorded while using a method in which hands were kept on the waist throughout the test. This value was similar to a value measured in physical education students.[32] A number of studies looked at vertical jump in players playing different positions. Only one study found a significant difference between guards and power forwards (48.9 – 4.9 cm vs 40.5 – 3.8 cm, respectively).[8] When converting the vertical jump values to anaerobic power, centres showed significantly higher anaerobic power compared with guards (108.5 – 12.7 vs 88.9 – 12.9 kg/m/sec, respectively).[9] In comparison, anaerobic power values were 120.7 – 4.0 kg/m/sec as obtained from the Margaria stair ª 2009 Adis Data Information BV. All rights reserved.
555
run,[23] and 67.67 kg/m/sec from the WanT.[6] The type of test selected by the researchers to assess anaerobic power greatly influenced the results in the different studies. Differences in skill levels were found to be related to vertical jump capability. A study that compared the best eight players in each playing position with the rest of the players in this position found that the best point guards had higher vertical jump values compared with the rest of the point guards (52.6 vs 44.8 cm, respectively). Similar results were found when comparing the best power forwards and the rest of the power forwards (50.5 vs 40.2 cm, respectively).[38] These results suggest that good jumping ability is associated with achieving success in basketball. This information can also be used for talent detection and early development processes in basketball, as well as for the establishment of conditioning programmes attempting to increase jumping height. For example, plyometric training has shown promising results in increasing jumping height.[41] 2.3.2 Male Players
One study found no differences in vertical jump and jumping power among players playing different positions,[14] and another found higher values of vertical jump height in guards and forwards compared with centres.[13] The latter study also found higher power in forwards and centres compared with guards. Differences in vertical jump among players at different skill levels revealed that the best players tend to jump higher compared with other players. However, in this study, only the difference between the best eight shooting guards and the rest of the shooting guards was found to be significant (68.6 vs 60.6 cm, respectively).[38] In a study of NBA players, no differences in power were found between guards and forwards.[15] The vertical jump values of basketball players were similar to those of football players, but vertical jump power values were higher in football players.[33] A comparison among basketball players, hockey players and soccer players revealed similar power values (120.4 – 10.7, 115.5 – 11.2 and 125.8 – 13.5 kg/m/sec, respectively).[21] Sports Med 2009; 39 (7)
Ziv & Lidor
556
Table II. Vertical jump values of male and female basketball players. Data shown as mean (SD) Study
Type of test
Position
Vertical jump (cm)
Guards (n = 7)
47.6 (4.9)
Females Bale[9] (1991)
CMJ
Forwards (n = 6)
47.2 (6.5)
Centres (n = 5)
47.6 (5.3)
Hakkinen[32] (1991)
Hands on waist; no arm swing allowed
All players (n = 9)
24.8 (2.5)
Hoare[38] (2000)
Step back with one foot, crouched positions, hands stretched back – bringing feet together, swing arms forward and leap
Guards (n = 62)
45.73 (NA)a
LaMonte et al.[6] (1999)
Riezebos et al.[23] (1983) Smith and Thomas
[8]
(1991)
CMJ
NA CMJ
Forwards (n = 44)
42.76 (NA)b
Centres (n = 19)
46.6 (4.8)
Guards (n = 18)
49.38 (6.20)
Forwards (n = 19)
49.43 (11.10)
Centres (n = 9)
43.51 (4.45)
All players (n = 20)
37.0 (1.1)d
Guards (n = 11)
48.9 (4.9)
Shooting forwards (n = 6)
44.5 (4.4)
Power forwards (n = 6)
40.5 (3.8)
Centres (n = 6)
42.0 (3.0)
Males Apostolidis et al.[18] (2004)
Hands on waist; no arm swing allowed
All players (n = 13)
40.1 (4.0)
Hakkinen[32] (1991)
Hands on waist; no arm swing allowed.
All players (n = 11)
43.9 (4.0)
Hoare[38] (2000)
Step back with one foot, crouched positions, hands stretched back – bringing feet together, swing arms forward and leap
Guards (n = 53)
63.3 (NA)a
Forwards (n = 56)
58.78 (NA)b
Hoffman et al.[34] (1996) Hoffman et al.
[39]
(2000)
Latin et al.[13] (1994)
Ostojic et al.[14] (2006)
Centres (n = 16)
57.9 (8.5)
CMJ
All players (n = 15)
68.5 (NA)c
Hands on waist; no arm swing allowed
All players (n = 9)
51.6 (6.9)
NA, data collected from surveys
Guards (n = 152)
73.4 (9.6)
Hands on waist; no arm swing allowed
a
Data combined for point guards and shooting guards by authors.
b
Data combined for small forwards and power forwards by authors.
d
Standard error of measurement on brackets.
c
Mean of four consecutive seasons (calculated by authors).
Forwards (n = 124)
71.4 (10.4)
Centres (n = 73)
66.8 (10.7)
Guards (n = 20)
59.7 (9.6)
Forwards (n = 20)
57.8 (6.5)
Centres (n = 20)
54.6 (6.9)
CMJ = counter movement jump (squatting with hands stretched backwards and leaping with hands swinging forward); NA = data not available.
When examining power output relative to bodyweight, similar results to those of the vertical jump were observed. Three studies used a 30-second all-out test to measure power output, ª 2009 Adis Data Information BV. All rights reserved.
two studies used the WanT,[18,39] and the third used a comparable test using an electromagnetically braked cycle ergometer.[17] Peak power outputs ranged from 10.7 – 1.3 to 14.1 – 1.4 W/kg. No Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
differences in peak power were indicated among players playing different positions, or among skill levels, although percentage fatigue was higher in the high-skilled than the low-skilled players (63.3% – 13.8 vs 54.1% – 11.1, respectively).[17] It can be assumed that the high-skilled players were more motivated to perform the test to the best of their ability. When examining changes in anaerobic capacity across the entire season, one study found lower anaerobic power values (as measured with an anaerobic step test) before pre-season began compared with 5 weeks after the beginning of pre-season.[28] No differences in power were found between 5 weeks after pre-season to the end of the season. Lastly, the comparison of basketball players with players in volleyball, handball, rugby and soccer resulted in similar relative peak power outputs (11.0 – 0.81, 11.2 – 0.64, 11.2 – 0.8, 10.9 – 0.59 and 10.6 – 0.68, respectively).[40] The large number of anaerobic power tests available for the strength and conditioning coach can be overwhelming. We suggest that the coach be aware of the concept of specificity in training and testing. For example, choosing anaerobic power tests that mimic actual game situations is most beneficial. In this respect, vertical jump tests are probably more useful for basketball players than the WanT. Choosing a specific vertical jump test is also important. Since arm swing is used during a basketball game, a counter movement jump with arm swing allowed would be preferable to a vertical jump test, in which arms are held behind the back or at the waist. 2.4 Agility and Speed
Agility and speed are integral aspects of almost every defensive and offensive manoeuvre performed by basketball players in practices and games. Only a few studies looked at these abilities: three in female[8,23,38] and five in male[13,18,33,34,38] players. 2.4.1 Female Players
Two studies looked at speed among players playing different positions.[8,38] When three conª 2009 Adis Data Information BV. All rights reserved.
557
secutive ‘suicide’ sprints (each consisting of a total of 143.3 m of running with seven reversals of direction) were performed, no differences were found among players in the first two sprints.[8] However, in the third sprint, guards were faster than centres. It was explained that fatigue probably affected the centres more than the guards. In another study,[38] point guards were quicker in an agility test compared with power forwards and centres. In addition, guards had better results in suicide sprints compared with power forwards. When achievements of players at different skill levels were compared, the best eight point guards had better results in 5 m, 10 m and 20 m suicide runs than the rest of the point guards participating in the study. On suicide sprints, the best eight small forwards were faster than the rest of the small forwards.[38] An analysis of the relationship between agility/speed and performance suggests that agility and speed are important characteristics for basketball players, and therefore including agility and speed drills in the conditioning programme is essential. Indeed, when the relationship between achievements in an agility run test and on-court performances – as evaluated by five experienced basketball coaches during game situations – was looked at, a moderate correlation was found (r = 0.4; p < 0.05).[23] 2.4.2 Male Players
A comparison of agility and speed among players playing different positions indicated conflicting results. In one study,[13] guards were faster than centres in both the 30-yard (27.4 m) dash and the 40-yard (36.6 m) dash. However, no differences were indicated among the players in an agility test. In another study,[38] point guards were faster than forwards and centres in an agility test; however, no differences were found among these players in 5 m, 10 m and 20 m runs. Since basketball is an intermittent sport and players have time to rest between bouts of intense activity, it is useful to examine the effects of the recovery mode – passive or active – between bouts on performance. One recent study[42] found that during a repeated-sprint test, passive recovery was associated with lower fatigue index and lower total sprint time. It was suggested that coaches Sports Med 2009; 39 (7)
Ziv & Lidor
558
should advise players to avoid unnecessary activity during the game, and should substitute players frequently when game intensity is high and little game interruption occurs (e.g. when there are few free throws or ball out of play). Comparison between the best eight players and the rest of the players indicated better suicide sprint times in the best eight shooting guards compared with the rest of the shooting guards. A comparison between basketball and football players revealed no differences in the 40-yard (36.6 m) dash.[33] The conflicting results obtained in the agility and speed tests are difficult to explain. It is assumed that the players would achieve similarly in both agility and speed tests. Future research is definitely needed on agility and speed in basketball players. Finally, correlational studies found that various field tests – such as control dribble, speed dribble, shuttle run and dribble shuttle run – were negatively correlated with %BF. However, these tests were positively correlated with peak power as measured by the WanT.[18] The obtained correlations were moderate, in the range 0.5–0.7. In addition, Hoffman et al.[34] found moderate correlations between achievements in speed and agility tests and on-court playing time. Agility and speed tests, like other fitness or physiological tests, should be in line with the concept of specificity. As Hoffman and Maresh suggested,[43] 30-yard (27.4 m) sprint tests are more specific to basketball players than 40-yard (36.6 m) sprints, as the court length is approximately 30 yards (27.4 m). It is up to the strength and conditioning coach to find the best agility and speed tests for the individual players in his or her team. 3. On-Court Performances Physiological variables such as heart rate and blood lactate can be measured not only under sterile laboratory conditions but also under field and more authentic conditions, namely actual games. Information on patterns of movements and actions performed by basketball players during the game can be also collected and analysed. In a number of studies, a time-motion ª 2009 Adis Data Information BV. All rights reserved.
analysis was used in order to quantify the number and types of movements performed by the players during a game. This analysis enabled the researchers to assess certain levels of the players’ workload during different phases of the game. 3.1 Female Players
Four studies examining on-court physiological demands of female players[11,20,22,44] were identified. Rodriguez-Alonso et al.[11] found higher blood lactate levels in guards compared with forwards and centres; however, no differences were indicated between national and international players. In addition, lower blood lactate levels were found when measured during practice games. The values of heart rates followed a similar pattern among players playing different positions, with guards showing higher heart rates during a game compared with forwards and centres. Unlike blood lactate, heart rates during international level play were higher than during national level play. McArdle et al.[22] showed mean heart rates of 81–95% of HRmax during an actual game. These values represent moderate to heavy workloads during a game. The authors also estimated oxygen consumption and average caloric cost of playing from heart rates, and suggested that players utilize 7.1–11.8 kcal/min. Only one study on time-motion analysis in female basketball players was found.[20] In this study, six female Division II basketball players participated in six practice games of approximately 20 minutes each (four quarters of 5 minutes with a 1-minute rest in between). In each game, one player’s physiological variables were measured using a portable metabolic system. For the purpose of this study, four movements were defined: standing, walking, running and jumping. On average, the players spent 1.6 – 0.9 minutes standing, 10.6 – 0.3 minutes walking, 6.2 – 0.7 minutes running and 0.3 – 0.1 minutes jumping. In addition, players spent approximately 34% of the time in running and jumping and other active movements. This study was also the first to measure oxygen consumption during a basketball game. The Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
players had an average oxygen consumption of 33.4 – 4.0 mLO 2/kg/min, which was 66.7 – 7.5% . of their VO2max.[20] Oxygen consumption remained. relatively the same throughout the game. These VO2 values are higher than those estimated in a recent compendium of physical activities (i.e. 28 mLO2/kg/min for a game of basketball),[45] suggesting greater utilization of aerobic metabolism than previously expected.[20] Blood lactate levels were 3.2 – 0.9 mmol/L measured every 5 minutes and were not changed throughout the game. It is difficult to establish whether these values represent a greater contribution from aerobic or anaerobic metabolism. Beam and Merill[44] recorded heart rates of female collegiate players in real game conditions. Results showed that players spent 61.8% of playing time with heart rates >90% of HRmax and 30.4% of the time with heart rates >95% of HRmax. This suggests that the intensity of female collegiate basketball is high enough to require large contributions from anaerobic metabolic pathways. 3.2 Male Players 3.2.1 Time-Motion Analysis
Only four time-motion analysis studies on male basketball players[19,20,46,47] were found. This is a low number for this kind of study, compared, for example, with another popular ball game – soccer.[48-55] A time-motion analysis was conducted by Ben Abdelkrim et al.[19] of elite under-19-year-old male basketball players. In this study, the authors defined nine specific movements, and the average total number of movements performed by the players during the game was 1050 – 51. McInnes et al.[46] defined eight specific movements (e.g. stand/walk, jog, run and stride/sprint), and examined their frequency and duration during a basketball game. The mean frequency of all movements was 997 – 183 and the mean duration of all movements was <3 seconds. The mean frequency of movements is smaller than that reported by Ben Abdelkrim et al.,[19] since these researchers defined eight types of movements (e.g. standing still, walking, jogging, and running) compared with nine in the study by Ben Abdelkrim et al.[19] ª 2009 Adis Data Information BV. All rights reserved.
559
In the study by McInnes et al.,[46] there was a change from one movement to another every 2 seconds, on average. This finding suggests that basketball is a game in which changes from one type of action to another are frequent, and hence agility and speed are important. Unlike the two previous studies, a third study on time-motion analysis was conducted during practice games of 20 minutes each of six male Division II college players.[20] The researchers defined only four movements for the analysis, and thus data cannot be compared with the previous two studies. It was found in this study that the players spent 1.7 – 0.6 minutes standing, 10.4 – 0.8 minutes walking, 5.8 – 0.8 minutes running and 0.3 – 0.1 minutes jumping. In addition, players spent approximately 34% of the time in active movements such as running and jumping. When examining the number of movements performed by players playing different positions, it was found that guards engaged in more movements compared with forwards and centres (1103 – 32 vs 1022 – 45 and 1026 – 27, respectively).[19] A number of studies investigated how frequently players engage in high-intensity movements during an actual game. Ben Abdelkrim et al.[19] found that guards and forwards spent a higher percentage of time performing highintensity movements compared with centres (17.1%, 16.6% vs 14.7%, respectively). Miller and Bartlett[47] found that centres spent more time stationary compared with guards and centres (32.8% vs 27.8% and 26.9%, respectively). It was also indicated by these authors that forwards spent the greatest amount of time running (18.6% of on-court time). In another study it was found that high-intensity movements were performed every 21 seconds (on average) during the actual time of play when the clock was running.[46] However, only 5% of stride/sprints performed by the players lasted more than 4 seconds, and therefore it seemed that the highest intensity sprints consisted of quick acceleration and deceleration without developing a full speed. As the authors suggested, a great deal of energy expenditure was associated with the need to overcome body inertia.[46] Altogether, only 15% of live time was spent on high-intensity movements. Sports Med 2009; 39 (7)
Ziv & Lidor
560
Nevertheless, 65% of live time was spent performing movements of greater intensity than walking. The changes in high-intensity movements as the game progressed were also examined in studies using a time-motion analysis. While McInnes et al.[46] found no differences in movement characteristics across quarters of play, Ben Abdelkrim et al.[19] reported that the percentage of highintensity movements was reduced from the first quarter to the second, and from the third to the fourth. This inconsistency might be related to a number of factors, among them: (i) the different fitness levels of the players who took part in these studies; (ii) the different training programmes they engaged in; (iii) the changes in the rules introduced by FIBA in 2000; and (iv) the different playing styles used throughout the quarters. It . was also found that VO2max values in the under19-year-old players were lower than those obtained from the adult players (approximately 53 and 60 mLO2/kg/min, respectively). When comparing performances of high-intensity movements in players playing different positions, it was found that guards engaged in more moderate-intensity specific movements than forwards and centres, while there were no differences among positions in low- and high-intensity movements.[19] Guards performed more sprints (mean – SD 67 – 5) than forwards (56 – 5) and centres (43 – 4), and forwards performed more sprints than centres. This information should be used by strength and conditioning coaches when planning training programmes for guards. Total distance covered by players throughout an actual game was not measured in either of the above studies. It was argued that since various movements in basketball are executed in a relatively small space (e.g. blocking, guarding and rebounding), the measurement of total distance could underestimate the physiological demands of the game.[46] However, knowing the total walking/jogging/running distance players cover during the game would probably help basketball coaches and strength and conditioning coaches improve the endurance regime of their training programmes. Only one study examined oxygen consumption during a practice game.[20] In this study, the ª 2009 Adis Data Information BV. All rights reserved.
players had an average oxygen consumption of was 64.7 – 7.0% 36.9 – 2.6 .mLO2/kg/min, which . of their VO2max. These VO2 values are higher than those estimated in a recent compendium of physical activities, i.e. 28 mLO2/kg/min for a game of basketball,[45] postulating a greater utilization of aerobic metabolism than previously expected.[20] 3.2.2 Heart Rate
Although heart rate can vary greatly among individuals, it has been considered a good predictor of exercise intensity. In one study, mean – SD HR was 169.3 – 4.5 beats/min during team scrimmage.[20] Expressed as a percentage of HRmax (using . values of peak HR that were reached during a VO2max test for those players) reveals an exercise intensity of approximately 88% of HRmax. In a study of 20 young basketball players (mean – SD age 16.8 – 2 years) during a 20-minute practice game, values of 86.2 – 5.3% and 86.7 – 4.3% of HRmax were recorded in the first and second halves, respectively.[56] Similarly, in a study by McInnes et al.[46] of male basketball players, it was observed that players had a HR >85% of HRmax in 75% of their playing time. Such high heart rates are usually associated with high intensity. However, it was indicated in the same study that high-intensity movements were performed during only 15% of the players’ playing time. The authors explained this discrepancy by suggesting that a variety of high-intensity movements – such as maintaining a position against physical resistance, passing, rebounding and shooting – were not measured in this study. A similar observation was made by Ben Abdelkrim et al.,[19] who compared heart rates among male players playing different positions. They found that guards had heart rates that were higher by 2–3 beats/min than forwards and centres. This difference suggested a slightly higher play intensity in guards, which is in line with the finding that guards were engaged in more moderate- and high-intensity movements. However, heart rate is also influenced by other variables, such as nutritional status,[11] environmental conditions,[11] psychological arousal,[19] anxiety[19] Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
and stoppage in game play.[19] Therefore, these data should be carefully interpreted. 3.2.3 Blood Lactate
Obtaining information on the metabolic pathways that are utilized during a basketball game – both aerobic and anaerobic – should also be of interest to the basketball coach and the strength and conditioning coach. Ben Abdelkrim et al.[19] reported that guards had higher blood lactate levels than centres (6.0 – 1.2 vs 4.9 – 1.1 mmol/L, respectively). Narazaki et al.[20] reported a mean value of 4.2 – 1.3 mmol/L in 20-minute practice games, while Castagna et al.[56] reported a mean of 3.72 – 1.39 in 20-min games of young basketball players (mean age 16.8 – 2 years). McInnes et al.[46] reported elevated lactate levels throughout a basketball game, with high variability among male players (mean maximum 8.5 – 3.1 mmol/L). The elevated lactate values suggested that glycolytic pathways made an important contribution to energy production during an actual game.[46] However, caution is warranted when interpreting blood lactate values. Blood lactate concentration is a snapshot of lactate turnover. Lactate is being produced by muscles working at high intensity, and at the same time it is being removed from those muscles to be used by other skeletal muscles, the cardiac muscle, or for gluconeogenesis in the liver. Therefore, the simple fact that blood lactate is elevated above resting levels does not tell us directly what percentage of energy comes from aerobic or anaerobic pathways. 4. Nutritional Strategies and Oxidative Stress It is widely accepted that maintaining proper nutrition is beneficial to athletic performance.[57] A number of studies examining the contribution of nutritional strategies to facilitating improved performance in basketball are reviewed. It is beyond the objectives and scope of our article to provide a more extensive review on nutritional strategies and oxidative stress in elite basketball players than that given here. Basketball players (female and male), as with other athletes, should maintain a positive energy ª 2009 Adis Data Information BV. All rights reserved.
561
balance and avoid low-energy intakes. Energy intake should be balanced between carbohydrates (55–58% of energy), proteins (12–15%) and fats (25–30%).[57] Special care should be taken in the vegetarian athlete, who might be at risk for low protein intake, as well as low micronutrients intake (specifically vitamin B12, zinc, iron and calcium). While most macronutrients and micronutrients can be supplied from foods in a balanced diet, additional supplements may be useful during an intense basketball season. Maintaining a positive and balanced energy intake can prove difficult between the practice sessions and games, and during long travels in a competitive basketball season. Schroder et al.[58] found that 32 (58%) of a sample of 55 basketball players in the First Spanish Basketball League reported using dietary supplements. Of those players, 81% used supplements on a daily basis, with multivitamins and vitamins being taken most frequently (50%). The authors suggested that the consumption of multivitamins might help in preventing temporary vitamin imbalances that may be caused by the frequency and timing of training sessions, travel and poor food selection. 4.1 Hydration
One nutrient that is often overlooked is water. Maintaining euhydration is important to aerobic performance, and it is suggested that a water deficit of 2% of bodyweight can lead to decreased performance.[59,60] A number of studies on hydration in basketball players suggest that dehydration is detrimental to performance.[61-63] In one study of eleven 17- to 28-year-old male basketball players, dehydration led to impaired vigilance-related attentional performance.[61] In another study of 17 male basketball players aged 17–28 years old, a progressive decline in basketball skills was associated with dehydration levels of 1–4% of bodyweight.[62] The threshold of water deficit at which overall performance deterioration became statistically significant was 2% of bodyweight.[62] Consumption of carbohydrate solutions (sports drinks) during intermittent exercise appears to improve sports performance.[64] In a study of 15 male adolescent players aged Sports Med 2009; 39 (7)
Ziv & Lidor
562
12–15 years,[63] it was found that 2% dehydration was associated with deterioration in basketball skill performance, and that euhydration with a 6% carbohydrate solution improved both shooting skills and on-court sprinting compared with euhydration with a placebo. Clearly, teaching basketball players to maintain hydration is important. This can be effectively accomplished by weighing each player before and after practices. In this manner, players can learn their individual sweat loss ratios and know how much fluid intake is required of them to maintain euhydration. 4.2 Oxidative Stress
Free radicals occur naturally in the body and can have negative effects on lipids, proteins and DNA oxidation. The antioxidant system alleviates these negative effects. When there is an imbalance between the production of free radicals and the antioxidant defence, oxidative stress occurs. Oxidative stress may be involved in the aging process, cell damage, some pathology, muscular fatigue and overtraining.[65] Exercise training increases the production of free radicals and the utilization of antioxidants. Therefore, proper nutrition is important in maintaining the antioxidants.[65] Since basketball players perform intense physical activity, their free radical production is likely to increase. Hence, it is important to supply the needed micronutrients that serve as antioxidants to alleviate the possible negative effects of the free radicals. When a balanced diet is not maintained, antioxidant supplements may be warranted. In one study,[66] researchers examined the effects of antioxidant supplements – a-tocopherol (vitamin E), b-carotene and ascorbic acid (vitamin C) – on exercise stress markers in 13 professional male basketball players (seven players in a supplement group, six players in a placebo group). After 32 days of treatment, elevated plasma levels of a-tocopherol and b-carotene were found in the supplement group but not in the placebo group. However, plasma levels of ascorbic acid were not elevated in the supplement ª 2009 Adis Data Information BV. All rights reserved.
group, while there was a significant decrease in the placebo group. The authors cautiously suggested the ascorbic acid levels remaining similar despite the supplementation might be explained by its use to scavenge free radicals and regenerate vitamin E. Importantly, lipid peroxide (lipid peroxidation is the degradation of lipids and can cause cell membrane damage) plasma concentration decreased in the supplement group by 27%, although the difference was not statistically significant (p < 0.09). The authors suggested that this may be related to a reduction in muscle cell damage during training.[66] Similar results were observed in another study,[67] which found an improvement in oxidative stress in elite male basketball players during a competitive season when antioxidant supplements were taken. A third study[68] found that a-tocopherol supplementation may reduce the DNA oxidation induced by training. In this study, total antioxidant status was higher after 1 month of supplementation. These studies suggest that basketball players may benefit from supplementing their diet with antioxidants. Interestingly, vegetarian athletes have higher antioxidant status for vitamin C, vitamin E and b-carotene compared with omnivores.[69] While the negative effects of free radicals do not usually affect performance, it is possible that they can lead to overtraining. This may be because muscular cell damage, which can be caused by free radicals, can reduce the metabolic capacities of muscle cells.[65] This speculation should be considered cautiously as there is no direct evidence to support it.[65] 5. Conditioning for Basketball A number of articles looked at conditioning for basketball.[43,70-74] Although an extensive review of the basketball conditioning literature is beyond the scope of our article, a few concepts are noteworthy. (For a detailed review of conditioning practices in basketball see Hoffman and Maresh.[43]) Conditioning practices for basketball players can be complex, as the players require good aerobic capacity, anaerobic power, speed, agility and strength. The limited time for conditioning Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
during the season means that the coaching staff working with the players need to decide what aspect of conditioning the strength and conditioning coach needs to concentrate on. Tavino et al.[28] suggested that during the offseason, players should use a combination of aerobic and anaerobic training to maintain fitness levels. During pre-season, athletes should concentrate more on developing anaerobic capacity, and during the season, high-intensity training should be done twice a week to maintain anaerobic capacity. In addition, weight training should be employed throughout the year, while during the season weight training should be carried out moderately twice a week to maintain strength.[28] Relevant information on conditioning programmes was obtained from NBA strength and conditioning coaches (NBA-SC) in a large survey of 20 NBA-SC where it was found that all worked with their players on flexibility, speed development, plyometrics and strength/power development. Much variation was seen in the types of drills, frequency, duration and intensity of training.[73] It was also found that most coaches (90%) divided their programmes into periods by taking into account the specific needs of their players in different phases of the season. NBA-SC also assessed the fitness level of their players. More coaches reported testing for aerobic capacity (n = 12) than anaerobic capacity (n = 10). It is unclear why less than half the coaches tested for anaerobic capacity, as it is clearly of great importance to the game. A time-motion analysis can be used by strength and conditioning coaches when planning conditioning programmes for their elite players. Taylor,[74] for example, suggested that four videotapes of game performance should be chosen for analysis for each player: (a) the best game of year; (b) the worst game of year; (c) the game with fewest fouls; and (d) a post-season game. After the analysis performed by the coach, he or she can plan workouts that simulate intensity and rest periods in order to mimic what players do in actual games. Including such workouts in the year-round conditioning programme can improve the physical preparation of the players for practices and games. ª 2009 Adis Data Information BV. All rights reserved.
563
Time-motion analysis can also be used in order to establish specific testing protocols for basketball players. This was the objective of Castagna et al.,[42] who decided on a repeatedsprint test protocol based on a time-motion analysis. They found that 93% of sprint stride sequences included no more than ten consecutive bouts and that sprint lengths were 5–32 m. Based on these data, they decided on a basketballspecific repeated-sprint test of ten shuttle run sprints of 15 m. As Hoffman et al.[75] indicated, once an aerobic base has been established, further increases in aerobic capacity may not increase performance in basketball players. Therefore, a maintenance programme consisting of running three times per week for a duration of 30–40 minutes may be sufficient. However, there is reason to believe that an aerobic training programme targeted at increasing the anaerobic threshold is beneficial as well. According to a recent study, the mean oxygen . is 66.7% of . consumption during game play VO2max in female and 64.7% of VO2max in male basketball players.[20] These values are probably in the vicinity of the anaerobic threshold of these players and perhaps a bit higher, although this was not measured in the study. That is to say that the players worked at an intensity that was at or slightly above their anaerobic threshold. Increasing .the anaerobic threshold to a realistic 70% of VO2max will allow players to use more aerobic metabolic pathways, which can lead to decreased fatigue during games. However, this recommendation should be considered with caution as it is inferred from only one study of six female and six male players. 5.1 Hormonal Status and the Overtraining Syndrome
One aspect of conditioning that should not be overlooked is overtraining. Peak performance can be achieved by the right combination of volume and intensity of training, as well as by providing the player with adequate resting periods in between.[76] If volume and intensity are too high, and if not enough recovery time is given to the player, overtraining can occur. The Sports Med 2009; 39 (7)
564
overtraining syndrome is characterized by diminished performance, increased fatigue and stress.[77] It has been suggested that disturbances of several hormones (e.g. growth hormone, cortisol,[77] hypothalamopituitary dysregulation[78]) can be reliable markers of overtraining. In addition, biochemical markers such as creatine kinase and urea can be indicative of muscle damage.[66] However, one study suggested that their validity in indicating overtraining may be overestimated.[78] It should be noted that there are no definite diagnostic criteria for the overtraining syndrome.[79] It should also be noted that the ratio between testosterone and cortisol represents the balance between anabolic and catabolic processes, and is also likely to represent the physiological strain of training.[77] One study was found that examined hormonal and biochemical changes in ten male basketball players participating in a 4-week training camp.[76] No difference in testosterone and luteinizing hormone levels in over 4 weeks of training was indicated. The training camp was scheduled 1 month after the end of the 9-month regular season. Plasma cortisol increased significantly in week 4 of the training programme, but remained within the normal range. The ratio between testosterone and cortisol was decreased by 22% in week 4, although this finding was not statistically significant. Between week 1 and week 2 of the training programme, creatine kinase levels increased by 60%; however, this was not statistically significant. This finding probably represented local muscle trauma only. It was concluded that a 4-week training camp for elite basketball players did not appear to cause any disturbance to the hormonal or biochemical profile of the players. Since it can be difficult to notice decrements in performance in basketball players, unlike sports like swimming or track and field in which measures of performance are quite clear,[76] more studies of hormonal and biochemical markers of overtraining that are evaluated over an entire basketball season are warranted. Such objective markers of overtraining can help coaches to start tapering early enough to prevent overtraining from developing. ª 2009 Adis Data Information BV. All rights reserved.
Ziv & Lidor
6. Five Limitations Observed from the Physical and Physiological Measurements The studies reviewed in this article provide useful information to both researchers and practitioners on various physical and physiological characteristics of female and male basketball players. Among these characteristics are height, mass, aerobic and anaerobic capacities, strength and agility. However, five limitations associated with the testing protocols used in the reviewed studies are presented here. (a) Lack of a longitudinal approach. In the majority of the reviewed studies the physical and physiological tests were given to the players only once. No replicated measurements across different periods of time were performed. In order to systematically examine characteristics of elite performers (e.g. basketball players), a longitudinal approach should be used as well. In this approach, one group of performers is observed over a long period of time,[80] enabling the researchers to collect data on a variety of dependent variables, as well as to study developmental perspectives of the observed group. From an expert theory perspective, it has been established[81] that a period of at least 10 years is required to achieve expertise in sport, as well as in other domains such as art, music and science. Therefore, it would be useful for researchers and practitioners alike to obtain information on the physical and physiological characteristics of elite players during different periods of time across the season/s, and among different groups of skill level and age. This information would result in improving the ability of coaches to compare achievements among players as well as to plan more effectively training programmes for elite basketball players. (b) Lack of tests performed under physical exertion conditions. The physical and physiological tests used in the reviewed studies were performed in a rested state, i.e. the players performed when they felt ready, according to the protocols of the tests. Fatigue primarily affects the central processes that take place between information receipt and the initiation of a movement.[82] In this Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
respect, it was shown that fatigue can influence certain mechanisms as they operate from the input of information to the output. Moderately high fatigue will impair performances requiring strength, endurance and rapid movements. Therefore, it is of particular interest to the coach to assess his or her players’ ability not only under rested conditions, but also under physical exertion conditions that reflect what is required of them in actual games. (c) Lack of studies using a time-motion analysis. The majority of the studies described in our review were conducted under laboratory conditions and sterile settings, where the players were instructed to perform the tests individually. In only a few studies were data collected on physical and physiological performances of players during actual games. In order to plan effective strength and conditioning programmes for basketball players, more information should be gathered on what actions players actually perform during the game. A systematic analysis of the main actions demonstrated by the players during the game should be carefully made, and then, based on this analysis, field observations using a timemotion analysis should be conducted on players of different skill levels as well as on players playing different positions. In addition, studying the work/rest ratio during competitive games by recording live physiological measurements (via a portable metabolic system) can allow coaches to gather additional relevant information on the players, and plan better conditioning programmes accordingly. (d) Reported HR values can be difficult to interpret. These values should be interpreted as a percentage of HRmax; however, these interpretations will be accurate only if the actual HRmax is known for the players, since the estimation of HRmax from age is inaccurate. In addition, HR values should be reported for the total game time, including stoppages. Recording HR only in live time ignores important recovery information during rest periods such as time-outs and half-time. (e) The results of blood lactate concentration can also be difficult to interpret. As mentioned earlier (section 3.2.3), blood lactate concentration is a result of lactate rate of appearance and rate of ª 2009 Adis Data Information BV. All rights reserved.
565
disappearance. Blood lactate and muscle lactate can be different during intermittent exercise, and blood lactate is not necessarily a good predictor of muscle lactate.[83] Moreover, while some of the working muscles produce lactate, other muscles that work at lower intensities may actually be consumers of lactate as a substrate. It is suggested that measuring or estimating the anaerobic threshold of each player, followed by the measurements of muscle and blood lactate over several games, as Ben Abdelkrim et al.[19] suggested, can enhance our understanding of the workload and metabolic pathways being used during a game. Several test protocols are available for the measurement of the anaerobic threshold. Performing those tests and interpreting their results require a knowledgeable and experienced staff. (For a review of the concept of the anaerobic threshold and the methods of measurement, see Svedahl and MacIntosh.[84])
7. Practical Advice for Basketball and Strength and Conditioning Coaches We have three recommendations for the basketball coach and the strength and conditioning coach. (a) Training programmes should be planned for the athlete according to his or her playing position. Guards, forwards and centres have different physical and physiological characteristics. Although relevant information on training programmes for basketball players can be found in the literature,[43] it is suggested that more emphasis be placed on developing specific programmes for forwards, centres and, particularly, for guards. Ultimately, basketball and strength and conditioning coaches should plan their training programmes according to the unique characteristics of each player. (b) A careful selection of the physical and physiological tests should be made. As reported in our review, there are a large number of different tests assessing physical and physiological abilities in basketball players. Therefore, it is recommended to carefully select the test most Sports Med 2009; 39 (7)
Ziv & Lidor
566
appropriate for assessing abilities in female and male basketball players. It is also recommended that the same test be used for comparing achievements in a specific ability among the basketball players. Different tests yield different values, and thus the resulting comparison would not be effective. (c) Coaches should be aware of changes in the rules of the game. They should take any changes made in the rules of the game into account while planning their training programmes. Practically speaking, the physical preparation as part of the annual training programme should also help the player adjust to any new changes made in the rules of the game.
8. Conclusions This paper reviewed several issues related to basketball players and their performance. Despite the five limitations observed in the reviewed studies, three practical recommendations for basketball coaches were noted. We suggest that future research should concentrate on timemotion analyses, physiological demands during game-play, and the effects of fatigue on performance. Acknowledgements The authors would like to thank Dinah Olswang for her editorial assistance during the preparation of this manuscript. No sources of funding were used to assist in the preparation of this review, and the authors have no conflicts of interest that are directly relevant to the content of this review.
References 1. Lidor R, Blumenstein B, Tenenbaum G. Psychological aspects of training in European basketball: conceptualization, periodization, and planning. Sport Psychologist 2007; 21: 353-67 2. Bompa T. Periodization: the theory and methodology of training. 4th ed. Champaign (IL): Human Kinetics, 1999 3. Matveyev L. Fundamentals of sports training. Moscow: Progress, 1981 4. Harre D, editor. Principles of sports training: introduction to theory and methods of training. Berlin: Sportverlag, 1982
ª 2009 Adis Data Information BV. All rights reserved.
5. Carter JE, Ackland TR, Kerr DA, et al. Somatotype and size of elite female basketball players. J Sports Sci 2005 Oct; 23 (10): 1057-63 6. LaMonte MJ, McKinney JT, Quinn SM, et al. Comparison of physical and physiological variables for female college basketball players. J Strength Cond Res 1999; 13 (3): 264-70 7. Ackland TR, Schreiner AB, Kerr DA. Absolute size and proportionality characteristics of World Championship female basketball players. J Sports Sci 1997 Oct; 15 (5): 485-90 8. Smith HK, Thomas SG. Physiological characteristics of elite female basketball players. Can J Sport Sci 1991 Dec; 16 (4): 289-95 9. Bale P. Anthropometric body composition and performance variables of young elite female basketball players. J Sports Med Phys Fitness 1991 Jun; 31 (2): 173-7 10. Bayios IA, Bergeles NK, Apostolidis NG, et al. Anthropometric, body composition and somatotype differences of Greek elite female basketball, volleyball and handball players. J Sports Med Phys Fitness 2006 Jun; 46 (2): 271-80 11. Rodriguez-Alonso M, Fernandez-Garcia B, Perez-Landaluce J, et al. Blood lactate and heart rate during national and international women’s basketball. J Sports Med Phys Fitness 2003 Dec; 43 (4): 432-6 12. Cormery B, Marcil M, Bouvard M. Rule change incidence on physiological characteristics of elite basketball players: a 10-year-period investigation. Br J Sports Med 2008 Jan; 42 (1): 25-30 13. Latin RW, Berg K, Baechle T. Physical and performance characteristics of NCAA division I male basketball players. J Strength Cond Res 1994; 8 (4): 214-8 14. Ostojic SM, Mazic S, Dikic N. Profiling in basketball: physical and physiological characteristics of elite players. J Strength Cond Res 2006 Nov; 20 (4): 740-4 15. Parr RB, Hoover R, Wilmore JH, et al. Professional basketball players: athletic profiles. Physician Sportsmed 1978 Apr; 6 (4): 77-9; 82-4 16. Viviani F. The somatotype of medium class Italian basketball players. J Sports Med Phys Fitness 1994 Mar; 34 (1): 70-5 17. Sallet P, Perrier D, Ferret JM, et al. Physiological differences in professional basketball players as a function of playing position and level of play. J Sports Med Phys Fitness 2005 Sep; 45 (3): 291-4 18. Apostolidis N, Nassis GP, Bolatoglou T, et al. Physiological and technical characteristics of elite young basketball players. J Sports Med Phys Fitness 2004 Jun; 44 (2): 157-63 19. Ben Abdelkrim N, El Fazaa S, El Ati J. Time-motion analysis and physiological data of elite under-19-year-old basketball players during competition. Br J Sports Med 2007 Feb; 41 (2): 69-75 20. Narazaki K, Berg K, Stergiou N, et al. Physiological demands of competitive basketball. Scand J Med Sci Sports. Epub 2008 Apr 21. Withers RT, Roberts GD, Davies GJ. The maximum aerobic power, anaerobic power and body composition of South Australian male representatives in athletics, basketball, field hockey and soccer. J Sports Med Phys Fitness 1977 Dec; 17 (4): 391-400
Sports Med 2009; 39 (7)
Attributes of Female and Male Basketball Players
22. McArdle WD, Magel JR, Kyvallos LC. Aerobic capacity, heart rate and estimated energy cost during women’s competitive basketball. Res Q 1971 May; 42 (2): 178-86 23. Riezebos ML, Paterson DH, Hall CR, et al. Relationship of selected variables to performance in women’s basketball. Can J Appl Sport Sci 1983 Mar; 8 (1): 34-40 24. Vaccaro P, Clarke DH, Wrenn JP. Physiological profiles of elite women basketball players. J Sports Med Phys Fitness 1979 Mar; 19 (1): 45-54 25. Caterisano A, Patrick BT, Edenfield WL, et al. The effects of a basketball season on aerobic and strength parameters among college men: starters versus reserves. J Strength Cond Res 1997; 11 (1): 21-4 26. Gocentas A, Landor A, Andziulis A. Dependence of intensity of specific basketball exercise from aerobic capacity. Pap Anthropol XIII 2004; 9-17 27. Parnat J, Viru A, Savi T, et al. Indices of aerobic work capacity and cardio-vascular response during exercise in athletes specializing in different events. J Sports Med Phys Fitness 1975 Jun; 15 (2): 100-5 28. Tavino LP, Bowers CJ, Archer CB. Effects of basketball on aerobic capacity, anaerobic capacity, and body composition of male college players. J Strength Cond Res 1995; 9 (2): 75-7 29. Vaccaro P, Wrenn JP, Clarke DH. Selected aspects of pulmonary function and maximal oxygen uptake of elite college basketball players. J Sports Med Phys Fitness 1980 Mar; 20 (1): 103-8 . 30. Castagna C, D’Ottavio S, Manzi V, et al. HR and VO2 responses during basketball drills [abstract]. In: Dikic N, Zivanic S, Ostojic S, et al., editors. Book of abstracts of the 10th Annual Congress of the European College of Sport Science 2005. Belgrade: 2005: 160 31. Whaley MH, Brubaker PH, Otto RM, editors. ACSM’s guidelines for exercise testing and prescription. 7th ed. Philadelphia (PA): American College of Sports Medicine, 2006 32. Hakkinen K. Force production characteristics of leg extensor, trunk flexor and extensor muscles in male and female basketball players. J Sports Med Phys Fitness 1991 Sep; 31 (3): 325-31 33. Berg K, Latin RW. Comparison of physical and performance characteristics of NCAA division I basketball and football players. J Strength Cond Res 1995; 14 (1): 22-6 34. Hoffman JR, Tenenbaum G, Maresh CM, et al. Relationship between athletic performance tests and playing time in elite college basketball players. J Strength Cond Res 1996; 10 (2): 67-71 35. Theoharopoulos A, Tsitskaris G, Nikopoulou M, et al. Knee strength of professional basketball players. J Strength Cond Res 2000; 14 (4): 453-63 36. Croisier JL. Factors associated with recurrent hamstring injuries. Sports Med 2004; 34 (10): 681-95 37. Knapik JJ, Bauman CL, Jones BH, et al. Preseason strength and flexibility imbalances associated with athletic injuries in female collegiate athletes. Am J Sports Med 1991 JanFeb; 19 (1): 76-81 38. Hoare DG. Predicting success in junior elite basketball players: the contribution of anthropometric and physio-
ª 2009 Adis Data Information BV. All rights reserved.
567
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
logical attributes. J Sci Med Sport 2000 Dec; 3 (4): 91-405 Hoffman JR, Epstein S, Einbinder M, et al. A comparison between the Wingate Anaerobic Power Test to both vertical jump and line drill tests in basketball players. J Strength Cond Res 2000; 14 (3): 261-4 Kalinski MI, Norkowski H, Kerner MS, et al. Anaerobic power characteristics of elite athletes in national level teamsport games. Eur J Sport Sci 2002; 2 (3): 1-14 Markovic G. Does plyometric training improve vertical jump height? A meta-analytical review. Br J Sports Med 2007 Jun; 41 (6): 349-55 Castagna C, Abt G, Manzi V, et al. Effect of recovery mode on repeated sprint ability in young basketball players. J Strength Cond Res 2008 May; 22 (3): 923-9 Hoffman JR, Maresh CM. Physiology of basketball. In: Garrett Jr WE, Kirkendall DT, editors. Exercise and sport science. Philadelphia (PA): Lippincott Williams & Wilkins, 2000: 733-44 Beam WC, Merrill TL. Analysis of heart rates recorded during female collegiate basketball [abstract]. Med Sci Sports Exerc 1994; 26 (5 Suppl.): S66 Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc 2000 Sep; 32 (9 Suppl.): S498-504 McInnes SE, Carlson JS, Jones CJ, et al. The physiological load imposed on basketball players during competition. J Sports Sci 1995 Oct; 13 (5): 387-97 Miller SA, Bartlett RM. Notational analysis of the physical demands of basketball [abstract]. J Sports Sci 1994; 12 (2): 181 Mohr M, Krustrup P, Bangsbo J. Match performance of high-standard soccer players with special reference to development of fatigue. J Sports Sci 2003 Jul; 21 (7): 519-28 Krustrup P, Mohr M, Ellingsgaard H, et al. Physical demands during an elite female soccer game: importance of training status. Med Sci Sports Exerc 2005 Jul; 37 (7): 1242-8 Burgess DJ, Naughton G, Norton KI. Profile of movement demands of national football players in Australia. J Sci Med Sport 2006 Aug; 9 (4): 334-41 Di Salvo V, Baron R, Tschan H, et al. Performance characteristics according to playing position in elite soccer. Int J Sports Med 2007 Mar; 28 (3): 222-7 Stroyer J, Hansen L, Klausen K. Physiological profile and activity pattern of young soccer players during match play. Med Sci Sports Exerc 2004 Jan; 36 (1): 168-74 Anastasiadis S, Anogeianaki A, Anogianakis G, et al. Real time estimation of physical activity and physiological performance reserves of players during a game of soccer. Stud Health Technol Inform 2004; 98: 13-5 Drust B, Atkinson G, Reilly T. Future perspectives in the evaluation of the physiological demands of soccer. Sports Med 2007; 37 (9): 783-805 Hill-Haas S, Coutts A, Rowsell G, et al. Variability of acute physiological responses and performance profiles of youth soccer players in small-sided games. J Sci Med Sport 2008; 11 (5): 487-90 Castagna C, Manzi V, Marini M, et al. Effect of playing basketball in young basketball players [abstract]. In:
Sports Med 2009; 39 (7)
Ziv & Lidor
568
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
Hoppeler H, Reilly T, Tsolakidis E, et al., editors. Book of abstracts of the 11th Annual Congress of the European College of Sport Science 2006. Lausanne: 2006: 325 Joint Position Statement: nutrition and athletic performance. American College of Sports Medicine, American Dietetic Association, and Dietitians of Canada. Med Sci Sports Exerc 2000 Dec; 32 (12): 2130-45 Schroder H, Navarro E, Mora J, et al. The type, amount, frequency and timing of dietary supplement use by elite players in the First Spanish Basketball League. J Sports Sci 2002 Apr; 20 (4): 353-8 Casa DJ, Clarkson PM, Roberts WO. American College of Sports Medicine roundtable on hydration and physical activity: consensus statements. Curr Sports Med Rep 2005 Jun; 4 (3): 115-27 Cheuvront SN, Carter 3rd R, Sawka MN. Fluid balance and endurance exercise performance. Curr Sports Med Rep 2003 Aug; 2 (4): 202-8 Baker LB, Conroy DE, Kenney WL. Dehydration impairs vigilance-related attention in male basketball players. Med Sci Sports Exerc 2007 Jun; 39 (6): 976-83 Baker LB, Dougherty KA, Chow M, et al. Progressive dehydration causes a progressive decline in basketball skill performance. Med Sci Sports Exerc 2007 Jul; 39 (7): 1114-23 Dougherty KA, Baker LB, Chow M, et al. Two percent dehydration impairs and six percent carbohydrate drink improves boys basketball skills. Med Sci Sports Exerc 2006 Sep; 38 (9): 1650-8 Coombes JS, Hamilton KL. The effectiveness of commercially available sports drinks. Sports Med 2000 Mar; 29 (3): 181-209 Finaud J, Lac G, Filaire E. Oxidative stress: relationship with exercise and training. Sports Med 2006; 36 (4): 327-58 Schroder H, Navarro E, Mora J, et al. Effects of alphatocopherol, beta-carotene and ascorbic acid on oxidative, hormonal and enzymatic exercise stress markers in habitual training activity of professional basketball players. Eur J Nutr 2001 Aug; 40 (4): 178-84 Schroder H, Navarro E, Tramullas A, et al. Nutrition antioxidant status and oxidative stress in professional basketball players: effects of a three compound antioxidative supplement. Int J Sports Med 2000 Feb; 21 (2): 146-50 Tsakiris S, Parthimos T, Tsakiris T, et al. Alpha-tocopherol supplementation reduces the elevated 8-hydroxy-2-deoxyguanosine blood levels induced by training in basketball players. Clin Chem Lab Med 2006; 44 (8): 1004-8 Venderley AM, Campbell WW. Vegetarian diets: nutritional considerations for athletes. Sports Med 2006; 36 (4): 293-305
ª 2009 Adis Data Information BV. All rights reserved.
70. Chandler J. Basketball: goals and activities for athletic conditioning in basketball. Strength Cond 1986; 8 (5): 52-5 71. Hilyer J, Hunter GR. Bridging the gap-practical application: a year-round strength development and conditioning program for men’s basketball. Strength Cond 1989; 11 (6): 16-9 72. Kroll WA. Conditioning for basketball. Strength Cond 1983; 5 (2): 24-6 73. Simenz CJ, Dugan CA, Ebben WP. Strength and conditioning practices of National Basketball Association strength and conditioning coaches. J Strength Cond Res 2005 Aug; 19 (3): 495-504 74. Taylor J. Basketball: applying time motion data to conditioning. Strength Cond 2003; 25 (2): 57-64 75. Hoffman JR, Epstein S, Einbinder M, et al. The influence of aerobic capacity on anaerobic performance and recovery indices in basketball players. J Strength Cond Res 1999; 13 (4): 407-11 76. Hoffman JR, Epstein S, Yarom Y, et al. Hormonal and biochemical changes in elite basketball players during a 4-week training camp. J Strength Cond Res 1999; 13 (3): 280-5 77. Urhausen A, Gabriel H, Kindermann W. Blood hormones as markers of training stress and overtraining. Sports Med 1995 Oct; 20 (4): 251-76 78. Urhausen A, Gabriel HH, Kindermann W. Impaired pituitary hormonal response to exhaustive exercise in overtrained endurance athletes. Med Sci Sports Exerc 1998 Mar; 30 (3): 407-14 79. Meeusen R, Duclos M, Gleeson M, et al. Prevention, diagnosis and treatment of the overtraining syndrome. Eur J Sport Sci 2006; 6 (1): 1-14 80. Thomas JR, Nelson JK. Research methods in physical activity. 5th ed. Champaign (IL): Human Kinetics, 2005 81. Ericsson KA. How the expert performance approach differs from traditional approaches to expertise in sports. In: Starkes JL, Ericsson KA, editors. Expert performance in sports-advances in research on sport expertise. Champaign (IL): Human Kinetics, 2003: 371-402 82. Pack M. Effects of four fatigue levels on performance and learning of novel dynamic balance skill. J Mot Behav 1974; 6: 191-7 83. Krustrup P, Mohr M, Steensberg A, et al. Muscle and blood metabolites during a soccer game: implications for sprint performance. Med Sci Sports Exerc 2006 Jun; 38 (6): 1165-74 84. Svedahl K, MacIntosh BR. Anaerobic threshold: the concept and methods of measurement. Can J Appl Physiol 2003 Apr; 28 (2): 299-323
Correspondence: Dr Ronnie Lidor, Associate Professor, The Zinman College of Physical Education and Sport Sciences, Wingate Institute, Netanya 42902, Israel. E-mail:
[email protected]
Sports Med 2009; 39 (7)
Sports Med 2009; 39 (7): 569-590 0112-1642/09/0007-0569/$49.95/0
REVIEW ARTICLE
ª 2009 Adis Data Information BV. All rights reserved.
Shoulder Muscle Recruitment Patterns and Related Biomechanics during Upper Extremity Sports Rafael F. Escamilla1, 2 and James R. Andrews1 1 Andrews-Paulos Research and Education Institute, Gulf Breeze, Florida, USA 2 Department of Physical Therapy, California State University, Sacramento, California, USA
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Shoulder Electromyography (EMG) during the Overhead Baseball Pitch . . . . . . . . . . . . . . . . . . . . . . . 1.1 Wind-Up Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Stride Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Arm Cocking Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Arm Acceleration Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 Arm Deceleration Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Shoulder EMG during the Overhead American Football Throw. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Shoulder EMG during Windmill Softball Pitching. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Shoulder EMG during the Volleyball Serve and Spike . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Shoulder EMG during the Tennis Serve and Volley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Shoulder EMG during Baseball Batting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Shoulder EMG during the Golf Swing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract
569 571 572 572 573 574 576 577 578 580 583 585 586 588
Understanding when and how much shoulder muscles are active during upper extremity sports is helpful to physicians, therapists, trainers and coaches in providing appropriate treatment, training and rehabilitation protocols to these athletes. This review focuses on shoulder muscle activity (rotator cuff, deltoids, pectoralis major, latissimus dorsi, triceps and biceps brachii, and scapular muscles) during the baseball pitch, the American football throw, the windmill softball pitch, the volleyball serve and spike, the tennis serve and volley, baseball hitting, and the golf swing. Because shoulder electromyography (EMG) data are far more extensive for overhead throwing activities compared with non-throwing upper extremity sports, much of this review focuses on shoulder EMG during the overhead throwing motion. Throughout this review shoulder kinematic and kinetic data (when available) are integrated with shoulder EMG data to help better understand why certain muscles are active during different phases of an activity, what type of muscle action (eccentric or concentric) occurs, and to provide insight into the shoulder injury mechanism. Kinematic, kinetic and EMG data have been reported extensively during overhead throwing, such as baseball pitching and football passing. Because
Escamilla & Andrews
570
shoulder forces, torques and muscle activity are generally greatest during the arm cocking and arm deceleration phases of overhead throwing, it is believed that most shoulder injuries occur during these phases. During overhead throwing, high rotator cuff muscle activity is generated to help resist the high shoulder distractive forces »80–120% bodyweight during the arm cocking and deceleration phases. During arm cocking, peak rotator cuff activity is 49–99% of a maximum voluntary isometric contraction (MVIC) in baseball pitching and 41–67% MVIC in football throwing. During arm deceleration, peak rotator cuff activity is 37–84% MVIC in baseball pitching and 86–95% MVIC in football throwing. Peak rotator cuff activity is also high is the windmill softball pitch (75–93% MVIC), the volleyball serve and spike (54–71% MVIC), the tennis serve and volley (40–113% MVIC), baseball hitting (28–39% MVIC), and the golf swing (28–68% MVIC). Peak scapular muscle activity is also high during the arm cocking and arm deceleration phases of baseball pitching, with peak serratus anterior activity 69–106% MVIC, peak upper, middle and lower trapezius activity 51–78% MVIC, peak rhomboids activity 41–45% MVIC, and peak levator scapulae activity 33–72% MVIC. Moreover, peak serratus anterior activity was »60% MVIC during the windmill softball pitch, »75% MVIC during the tennis serve and forehand and backhand volley, »30–40% MVIC during baseball hitting, and »70% MVIC during the golf swing. In addition, during the golf swing, peak upper, middle and lower trapezius activity was 42–52% MVIC, peak rhomboids activity was »60% MVIC, and peak levator scapulae activity was »60% MVIC.
Electromyography (EMG) is the science of quantifying muscle activity. Several studies have reported shoulder muscle activity during a variety of upper extremity sports.[1-7] Understanding when and how much specific shoulder muscles are active during upper extremity sports is helpful to physicians, therapists, trainers and coaches in providing appropriate treatment, training and rehabilitation protocols to these athletes, as well as helping health professionals better understand the shoulder injury mechanism. When interpreting EMG data it should be emphasized that while the EMG amplitude does correlate reasonably well with muscle force for isometric contractions, it does not correlate well with muscle force as muscle contraction velocities increase, or during muscular fatigue (both of which occur in sport).[8] Nevertheless, EMG analyses are helpful in determining the timing and quantity of muscle activation throughout a given movement. This review focuses on shoulder muscle activity in upper extremity sports, specifically: baseball pitching, American football throwing, windmill ª 2009 Adis Data Information BV. All rights reserved.
softball pitching, the volleyball serve and spike, the tennis serve and volley, baseball hitting, and the golf swing. Most of the movements that occur in the aforementioned sports involve overhead throwing type movements. Shoulder EMG data in the literature are far more extensive for overhead throwing activities, such as baseball pitching, compared with other upper extremity sports that do not involve the overhead throwing motion, such as baseball hitting. Therefore, much of this review focuses on shoulder EMG during activities that involve the overhead throwing motion. To help better interpret the applicability and meaningfulness of shoulder EMG data, EMG data will be integrated with shoulder joint kinematics (linear and angular shoulder displacements, velocities and accelerations) and kinetics (shoulder forces and torques) when these data are available. In the literature, kinematic, kinetic and EMG measurements have been reported extensively in overhead throwing activities,[2,9-12] such as baseball pitching and football throwing, but these data are sparse in other upper extremity activities, such as Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
571
the volleyball serve and spike, the tennis serve and volley, baseball hitting, and the golf swing. Overhead throwing activities in particular are commonly associated with shoulder injuries.[13,14] When EMG is interpreted with shoulder kinematics and kinetics, it not only provides a better understanding of why certain muscles are active during different phases of an activity, but also provides information as to what type of muscle action (eccentric or concentric) is occurring, and insight into the shoulder injury mechanism. Although shoulder muscle activity is the primary focus of this review, shoulder injuries will be dis-
cussed briefly relative to joint loads, joint motions and muscle activity when these data are available. 1. Shoulder Electromyography (EMG) during the Overhead Baseball Pitch Shoulder muscle activity during baseball pitching has been examined extensively by Jobe and colleagues,[2,15-18] with their initial report published in 1983.[18] Using 56 healthy male college and professional pitchers, DiGiovine and colleagues[2] quantified shoulder muscle activity
Table I. Shoulder activity by muscle and phase during baseball pitchinga (adapted from DiGiovine et al.,[2] with permission) Muscles
No. of subjects
Phase wind-upb (% MVIC)
Upper trapezius
11
18 – 16
64 – 53
37 – 29
Middle trapezius
11
7–5
43 – 22
51 – 24
Lower trapezius
13
13 – 12
39 – 30
Serratus anterior (6th rib)
11
14 – 13
Serratus anterior (4th rib)
10
Rhomboids Levator scapulae
stridec (% MVIC)
arm cockingd (% MVIC)
arm acceleratione (% MVIC)
arm decelerationf (% MVIC)
follow-throughg (% MVIC)
69 – 31
53 – 22
14 – 12
71 – 32
35 – 17
15 – 14
38 – 29
76 – 55
78 – 33
25 – 15
44 – 35
69 – 32
60 – 53
51 – 30
32 – 18
20 – 20
40 – 22
106 – 56
50 – 46
34 – 7
41 – 24
11
7–8
35 – 24
41 – 26
71 – 35
45 – 28
14 – 20
11
6–5
35 – 14
72 – 54
76 – 28
33 – 16
14 – 13
Anterior deltoid
16
15 – 12
40 – 20
28 – 30
27 – 19
47 – 34
21 – 16
Middle deltoid
14
9–8
44 – 19
12 – 17
36 – 22
59 – 19
16 – 13
Posterior deltoid
18
6–5
42 – 26
28 – 27
68 – 66
60 – 28
13 – 11
Supraspinatus
16
13 – 12
60 – 31
49 – 29
51 – 46
39 – 43
10 – 9
Infraspinatus
16
11 – 9
30 – 18
74 – 34
31 – 28
37 – 20
20 – 16
Teres minor
12
5–6
23 – 15
71 – 42
54 – 50
84 – 52
25 – 21
Subscapularis (lower 3rd)
11
7–9
26 – 22
62 – 19
56 – 31
41 – 23
25 – 18
Subscapularis (upper 3rd)
11
7–8
37 – 26
99 – 55
115 – 82
60 – 36
16 – 15
Pectoralis major
14
6–6
11 – 13
56 – 27
54 – 24
29 – 18
31 – 21
Latissimus dorsi
13
12 – 10
33 – 33
50 – 37
88 – 53
59 – 35
24 – 18
Triceps brachii
13
4–6
17 – 17
37 – 32
89 – 40
54 – 23
22 – 18
Biceps brachii
18
8–9
22 – 14
26 – 20
20 – 16
44 – 32
16 – 14
Scapular
Glenohumeral
a
Data are given as means and standard deviations, and expressed for each muscle as a percentage of an MVIC.
b
From initial movement to maximum knee lift of stride leg.
c
From maximum knee lift of stride leg to when lead foot of stride leg initially contacts the ground.
d
From when lead foot of stride leg initially contacts the ground to maximum shoulder external rotation.
e
From maximum shoulder external rotation to ball release.
f
From ball release to maximum shoulder internal rotation.
g
From maximum shoulder internal rotation to maximum shoulder horizontal adduction.
MVIC = maximum voluntary isometric contraction.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Escamilla & Andrews
572
Knee up
Phases
Wind-up
Foot contact
Stride
Max ER
Arm cocking
Release
Arm acceleration
Arm deceleration
Max IR
Follow-through
Fig. 1. Pitching phases and key events (adapted from Fleisig et al.,[12] with permission). ER = external rotation; IR = internal rotation; max = maximum.
during baseball pitching (data summarized in table I). To help generalize phase comparisons in muscle activity from table I, 0–20% of a maximum voluntary isometric contraction (MVIC) is considered low muscle activity, 21–40% MVIC is considered moderate muscle activity, 41–60% MVIC is considered high muscle activity and >60% MVIC is considered very high muscle activity.[2] From these initial reports, the baseball pitch was divided into several phases, which later were slightly modified by Escamilla et al.[9] and Fleisig et al.[11] as the wind-up, stride, arm cocking, arm acceleration, arm deceleration and follow-through phases (figure 1). 1.1 Wind-Up Phase
Shoulder activity during the wind-up phase, which is from initial movement to maximum knee lift of stride leg (figure 1), is generally very low due to the slow movements that occur. From table I, it can be seen that the greatest activity is from the upper trapezius, serratus anterior and anterior deltoids. These muscles all contract concentrically to upwardly rotate and elevate the scapula and abduct the shoulder as the arm is initially brought overhead, and then contract eccentrically to control downward scapular rotation and shoulder adduction as the hands are lowered to approximately chest level. The rotator ª 2009 Adis Data Information BV. All rights reserved.
cuff muscles, which have a duel function as glenohumeral joint compressors and rotators, have their lowest activity during this phase. Because shoulder activity is low, it is not surprising that the shoulder forces and torques generated are also low;[9,11] consequently, very few, if any, shoulder injuries occur during this phase. 1.2 Stride Phase
There is a dramatic increase in shoulder activity during the stride phase (table I), which is from the end of the balance phase to when the lead foot of the stride leg initially contacts the ground (figure 1). During the stride the hands separate, the scapula upwardly rotates, elevates and retracts, and the shoulders abduct, externally rotate and horizontally abduct due to concentric activity from several muscles, including the deltoids, supraspinatus, infraspinatus, serratus anterior and upper trapezius. It is not surprising that there are many more muscles activated and to a higher degree during the stride compared with the windup phase. Interestingly, the supraspinatus has its highest activity during the stride phase as it works to not only abduct the shoulder but also help compress and stabilize the glenohumeral joint.[2] The deltoids exhibit high activity during this phase in order to initiate and maintain the shoulder in an abducted position.[2] Moreover, the trapezius Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
and serratus anterior have moderate to high activity, as they assist in stabilizing and properly positioning the scapula to minimize the risk of impingement as the arm abducts.[2] 1.3 Arm Cocking Phase
The arm cocking phase begins at lead foot contact and ends at maximum shoulder external rotation. During this phase the kinetic energy that is generated from the larger lower extremity and trunk segments is transferred up the body to the smaller upper extremity segments.[10,19,20] The pitching arm lags behind as the trunk rapidly rotates forward to face the hitter, generating a peak pelvis angular velocity around 600/sec occurring 0.03–0.05 sec after lead foot contact, followed by a peak upper torso angular velocity of nearly 1200/sec occurring 0.05–0.07 sec after lead foot contact.[10] Consequently, high to very high shoulder muscle activity is needed during this phase in order to keep the arm moving with the rapidly rotating trunk (table I), as well as control the resulting shoulder external rotation (table I), which peaks near 180.[10] Moderate activity is needed by the deltoids (table I) to maintain the shoulder at approximately 90 abduction throughout this phase.[10] Activity from the pectoralis major and anterior deltoid is needed during this phase to horizontally adduct the shoulder with a peak angular velocity of approximately 600/sec, from a position of approximately 20 of horizontal abduction at lead foot contact to a position of approximately 20 of horizontal adduction at maximum shoulder external rotation.[10] Moreover, a large compressive force of »80% bodyweight is generated by the trunk onto the arm at the shoulder to resist the large ‘centrifugal’ force that is generated as the arm rotates forward with the trunk.[11] The supraspinatus, infraspinatus, teres minor and subscapularis achieve high to very high activity (table I) to resist glenohumeral distraction and enhance glenohumeral stability. While it is widely accepted that strength and endurance in posterior shoulder musculature is very important during the arm deceleration phase to slow down the arm, posterior shoulder musª 2009 Adis Data Information BV. All rights reserved.
573
culature is also important during arm cocking. The posterior cuff muscles (infraspinatus and teres minor) and latissimus dorsi generate a posterior force to the humeral head that helps resist anterior humeral head translation, which may help unload the anterior capsule and anterior band of the inferior glenohumeral ligament.[11,15,21] The posterior cuff muscles (infraspinatus and teres minor) also contribute to the extreme range of shoulder external rotation that occurs during this phase. A peak shoulder internal rotation torque of 65–70 N m is generated near the time of maximum shoulder external rotation,[11,22] which implies that shoulder external rotation is progressively slowing down as maximum shoulder external rotation is approached. High to very high activity is generated by the shoulder internal rotators (pectoralis major, latissimus dorsi and subscapularis) [table I], which contract eccentrically during this phase to control the rate of shoulder external rotation.[2] The multiple functions of muscles are clearly illustrated during arm cocking. For example, the pectoralis major and subscapularis contract concentrically to horizontally adduct the shoulder and eccentrically to control shoulder external rotation. This duel function of these muscles helps maintain an appropriate length-tension relationship by simultaneously shortening and lengthening, which implies that these muscles may be maintaining a near constant length throughout this phase. Therefore, some muscles that have duel functions and simultaneous shortening and lengthening as the shoulder performs duel actions at the same time may in effect be contracting isometrically. The importance of scapular muscles during arm cocking is demonstrated in table I. High activity from these muscles is needed in order to stabilize the scapula and properly position the scapula in relation to the horizontally adducting and rotating shoulder. The scapular protractors are especially important during this phase in order to resist scapular retraction by contracting eccentrically and isometrically during the early part of this phase and cause scapular protraction by contracting concentrically during the latter
Sports Med 2009; 39 (7)
Escamilla & Andrews
574
part of this phase. The serratus anterior generates maximum activity during this phase. Scapular muscle imbalances may lead to abnormal scapular movement and position relative to the humerus, increasing injury risk. Because both the triceps brachii (long head) and biceps brachii (both heads) cross the shoulder, they both generate moderate activity during this phase in order to provide additional stabilization to the shoulder. In contrast to the moderate triceps activity reported by DiGiovine et al.[2] during arm cocking, Werner et al.[23] reported the highest triceps activity during arm cocking. Because elbow extensor torque peaks during this phase,[23,24] high eccentric contractions by the triceps brachii are needed to help control the rate of elbow flexion that occurs throughout the initial 80% of this phase.[10] High triceps activity is also needed to initiate and accelerate elbow extension, which occurs during the final 20% of this phase as the shoulder continues externally rotating.[10] Therefore, during arm cocking the triceps initially contract eccentrically to control elbow flexion early in the phase and concentrically to initiate elbow extension later in the phase. Gowan and colleagues[16] demonstrated that subscapularis activity is nearly twice as great in professional pitchers compared with amateur pitchers during this phase. In contrast, muscle activity from the pectoralis major, supraspinatus, serratus anterior and biceps brachii was »50% greater in amateur pitchers compared with professional pitchers. From these data, professional pitchers may exhibit better throwing efficiency thus requiring less muscular activity compared with amateurs. Glousman and colleagues[15] compared shoulder muscle activity between healthy pitchers with no shoulder pathologies to pitchers with chronic anterior shoulder instability due to anterior glenoid labral tears. Pitchers diagnosed with chronic anterior instability exhibited greater muscle activity from the biceps brachii and supraspinatus and less muscle activity from the pectoralis major, subscapularis and serratus anterior. Chronic anterior instability results in excessive stretch of the anterior capsular, which may stimulate mechanoª 2009 Adis Data Information BV. All rights reserved.
receptors within the capsule resulting in excitation in the biceps brachii and supraspinatus and inhibition in the pectoralis major, subscapularis and serratus anterior.[15] Increased activity from the biceps brachii and supraspinatus helps compensate for anterior shoulder instability, as these muscles enhance glenohumeral stability. Rodosky et al.[25] reported that as the humerus abducts and maximally externally rotates, the biceps long head enhances anterior stability of the glenohumeral joint and also decreases the stress placed on the inferior glenohumeral ligament. Decreased activity from the pectoralis major and subscapularis, which contract eccentrically to decelerate the externally rotating shoulder, may accentuate shoulder external rotation and increase the stress on the anterior capsule.[15] Decreased activity from the serratus anterior may cause the scapula to be abnormally positioned relative to the externally rotating and horizontally adducting humerus, and a deficiency in scapular upward rotation may decrease the subacromial space and increase the risk of impingement and rotator cuff pathology.[26] Interestingly, infraspinatus activity was lower in pitchers with chronic anterior shoulder instability compared with healthy pitchers.[16] During arm cocking, the infraspinatus not only helps externally rotate and compress the glenohumeral joint, but also may generate a small posterior force on the humeral head due to a slight posterior orientation of its fibres as they run from the inferior facet of the greater tubercle back to the infraspinous fossa. As previously mentioned, this posterior force on the humeral head helps resist anterior humeral head translation and unloads strain on the anterior capsule during arm cocking.[16] It is unclear whether chronic rotator cuff insufficiency results in shoulder instability, or whether chronic shoulder instability results in rotator cuff insufficiency due to excessive activity. 1.4 Arm Acceleration Phase
The arm acceleration phase begins at maximum shoulder external rotation and ends at ball release[10,11,22] (figure 1). Like the arm cocking phase, high to very high activity is generated from the glenohumeral and scapular muscles during Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
this phase in order to accelerate the arm forward (table I). Moderate activity is generated by the deltoids[2] to help produce a fairly constant shoulder abduction of approximately 90–100,[10] which is maintained regardless of throwing style (i.e. overhand, sidearm, etc.). The glenohumeral internal rotators (subscapularis, pectoralis major and latissimus dorsi) have their highest activity during this phase[2] (table I) as they contract concentrically to help generate a peak internal rotation angular velocity of approximately 6500/sec near ball release.[9] This rapid internal rotation, with a range of motion of approximately 80 from maximum external rotation to ball release, occurs in only 30–50 msec.[10,27] The very high activity from the subscapularis (115% MVIC) occurs in part to help generate this rapid motion, but it also functions as a steering muscle to maintain the humeral head in the glenoid. The teres minor, infraspinatus and supraspinatus also demonstrate moderate to high activity during this phase to help properly position the humeral head within the glenoid. With these rapid arm movements that are generated to accelerate the arm forward, it is not surprising that the scapular muscles also generate high activity,[2] which is needed to help maintain proper position of the glenoid relative to the rapidly moving humeral head. Strengthening scapular musculature is very important because poor position and movement of the scapula can increase the risk of impingement and other related injuries,[28] as well as reduce the optimal length-tension relationship of both scapular and glenohumeral musculature. Although DiGiovine et al.[2] reported that the triceps had their highest activity during this phase,[2] Werner et al.[23] reported relatively little triceps EMG during the arm acceleration phase. In addition, elbow extensor torque is very low during this phase compared with the arm cocking phase.[23,24] It should be re-emphasized that elbow extension initially begins during the arm cocking phase as the shoulder approaches maximum external rotation.[9] Kinetic energy that is transferred from the lower extremities and trunk to the arm is used to help generate a peak elbow extension angular velocity of approximately 2300/sec during this phase.[9] In fact, a conª 2009 Adis Data Information BV. All rights reserved.
575
centric contraction from the triceps brachii alone could not come close to generating this 2300/sec elbow extension angular velocity. This is supported by findings reported by Roberts,[29] who had found that subjects who threw with paralyzed triceps could obtained ball velocities >80% of the ball velocities obtained prior to the triceps being paralyzed. This is further supported by Toyoshima et al.,[20] who demonstrated normal throwing using the entire body generated almost twice the elbow extension angular velocity compared with extending the elbow by throwing without any lower extremity, trunk and shoulder movements. These authors concluded that during normal throwing the elbow is swung open like a ‘whip’, primarily due to linear and rotary contributions from the lower extremity, trunk and shoulder, and to a lesser extent from a concentric contraction of the triceps. Nevertheless, the triceps do help extend the elbow during this phase, as well as contribute to shoulder stabilization by the triceps long head. These findings illustrate the importance of lower extremity conditioning, because weak or fatigued lower extremity musculature during throwing may result in increased loading of the shoulder structures, such as the rotator cuff, glenoid labrum, and shoulder capsule and ligaments. Further research is needed to substantiate these hypotheses. Gowan and colleagues[16] demonstrated that rotator cuff and biceps brachii activity was 2–3 times higher in amateur pitchers compared with professional pitchers during this phase. In contrast, subscapularis, serratus anterior and latissimus dorsi activity was much greater in professional pitchers. These results imply that professional pitchers may better coordinate body segment movements to increase throwing efficiency. Enhanced throwing mechanics and efficiency may minimize glenohumeral instability during this phase, which may help explain why professional pitchers generate less rotator cuff and biceps activity, which are muscles that help resist glenohumeral joint distraction and enhance stability. Compared with healthy pitchers, pitchers with chronic anterior shoulder instability due to anterior labral injuries exhibit greater muscle activity from the biceps brachii, supraspinatus and Sports Med 2009; 39 (7)
Escamilla & Andrews
576
infraspinatus, and less muscle activity from the latissimus dorsi, subscapularis and serratus anterior.[15] The increased activity from rotator cuff and biceps musculature in pitchers with chronic anterior instability is needed in order to provide additional glenohumeral instability that is lacking in these pitchers due to a compromised anterior labrum. With shoulder internal rotation, the long biceps tendon is repositioned anteriorly at the shoulder, providing compressive and posterior forces to the humeral head, both of which enhance anterior stability. Therefore, throwers with chronic anterior instability activate their biceps to a greater extent (32% vs 12% MVIC), as well as their supraspinatus and infraspinatus (37% vs 13% MVIC), compared with asymptomatic throwers.[15] However, increased and excessive biceps activity due to anterior instability results in increased stress to the long biceps anchor at the superior labrum, which over time may result in superior labral pathology that is anterior to posterior in direction (SLAP lesions). In addition, chronic anterior shoulder instability inhibits normal contributions from the internal rotators and serratus anterior,[15] which may adversely affect throwing mechanics and efficiency, as well as increase shoulder injury risk. 1.5 Arm Deceleration Phase
The arm deceleration phase begins at ball release and ends at maximum shoulder internal rotation (figure 1).[10,11,22] Large loads are generated at the shoulders to slow down the forward acceleration of the arm. The purpose of this phase is to provide safety to the shoulder by dissipating the excess kinetic energy not transferred to the ball, thereby minimizing the risk of shoulder injury. Posterior shoulder musculature, such as the infraspinatus, teres minor and major, posterior deltoid and latissimus dorsi, contract eccentrically not only to decelerate horizontal adduction and internal rotation of the arm, but also help resist shoulder distraction and anterior subluxation forces. A shoulder compressive force slightly greater than bodyweight is generated to resist shoulder distraction, while a posterior shear force of 40–50% bodyweight is generated to resist ª 2009 Adis Data Information BV. All rights reserved.
shoulder anterior subluxation.[9,11] Consequently, high activity is generated by posterior shoulder musculature,[2] in particular the rotator cuff muscles. For example, the teres minor, which is a frequent source of isolated tenderness in pitchers, exhibits its maximum activity (84% MVIC) during this phase (table I). In addition, scapular muscles also exhibit high activity to control scapular elevation, protraction and rotation during this phase. For example, the lower trapezius – which generate a force on the scapula in the direction of depression, retraction and upward rotation – generated their highest activity during this phase (table I). High EMG activity from glenohumeral and scapular musculature illustrate the importance of strength and endurance training of the posterior musculature in the overhead throwing athlete. Weak or fatigued posterior musculature can lead to multiple injuries, such as tensile overload undersurface cuff tears, labral/biceps pathology, capsule injuries and internal impingement of the infraspinatus/ supraspinatus tendons on the posterosuperior glenoid labrum.[14] Compared with healthy pitchers, pitchers with chronic anterior shoulder instability exhibited less muscle activity from the pectoralis major, latissimus dorsi, subscapularis and serratus anterior, which is similar to what occurred in the arm cocking and acceleration phases.[15] However, muscle activities from the rotator cuff and biceps brachii are similar between healthy pitchers and pitchers with chronic anterior shoulder instability during this phase, which is in contrast to the greater rotator cuff and biceps brachii activity demonstrated in pitchers with chronic anterior shoulder instability during the arm cocking and acceleration phases.[15] This difference in muscle activity may partially be explained by the very high compressive forces that are needed during arm deceleration to resist shoulder distraction, which is a primary function of both the rotator cuff and biceps brachii. The biceps brachii generate their highest activity (44% MVIC) during arm deceleration (table I). The function of this muscle during this phase to 2-fold. Firstly, it must contract eccentrically along with other elbow flexors to help Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
decelerate the rapid elbow extension that peaks near 2300/sec during arm acceleration.[9] This is an important function because weakness or fatigue in the elbow flexors may result in elbow extension being decelerated by impingement of the olecranon in the olecranon fossa, which may lead to bone spurs and subsequent loose bodies within the elbow. Secondly, the biceps brachii works synergistically with the rotator cuff muscles to resist distraction and anterior subluxation at the glenohumeral joint. Interestingly, during arm deceleration biceps brachii activity is greater in amateur pitchers compared with professional pitchers,[16] which may imply that amateur pitchers employ a less efficient throwing pattern compared with professional pitchers. As previously mentioned, excessive activity from the long head of the biceps brachii may lead to superior labral pathology. 2. Shoulder EMG during the Overhead American Football Throw There is only one known study that has quantified muscle activity during the football throw.[3] Using 14 male recreational and college athletes,
577
Kelly et al.[3] quantified activity from nine glenohumeral muscles throughout throwing phases specific for football; their results are summarized in table II. The defined phases for football throwing (table II) are similar but slightly different to the defined phases for baseball pitching (table I). Early arm cocking in the football throw was similar to the stride phase in baseball, while late cocking in the football throw was the same as arm cocking in baseball. The acceleration phase was the same for both the football throw and the baseball pitch. The arm deceleration and follow-through phases in the baseball pitch were combined into a single arm deceleration/follow-through phase in the football throw. From table II, rotator cuff activity progressively increased in each phase of the football throwing, being least in the early cocking phase and peaking in the arm deceleration/followthrough phase. This is a slightly different pattern than the baseball pitch, where rotator cuff activity was generally greatest during either the arm cocking phase or the arm deceleration phase (table I). For both baseball pitching and football throwing, deltoid and biceps brachii activity were generally greatest during the arm deceleration
Table II. Shoulder activity by muscle and phase during the overhead football throwa (adapted from Kelly et al.,[3] with permission) Muscles
No. of subjects
Phase early cockingb (% MVIC)
late cockingc (% MVIC)
arm accelerationd (% MVIC)
arm deceleration and follow-throughe (% MVIC)
total throwf (% MVIC)
Supraspinatus
14
45 – 19
62 – 20
65 – 30
87 – 43
65 – 22
Infraspinatus
14
46 – 17
67 – 19
69 – 29
86 – 33
67 – 21
Subscapularis
14
24 – 15
41 – 21
81 – 34
95 – 65
60 – 28
Anterior deltoid
14
13 – 9
40 – 14
49 – 14
43 – 26
36 – 9
Middle deltoid
14
21 – 12
14 – 14
24 – 14
48 – 19
27 – 9
Posterior deltoid
14
11 – 6
11 – 15
32 – 22
53 – 25
27 – 11
Pectoralis major
14
12 – 14
51 – 38
86 – 33
79 – 54
57 – 27
Latissimus dorsi
14
7–3
18 – 9
65 – 30
72 – 42
40 – 12
Biceps brachii
14
12 – 7
12 – 10
11 – 9
20 – 18
14 – 9
a
Data are given as means and standard deviations, and expressed for each muscle as a percentage of a MVIC.
b
From rear foot plant to maximum shoulder abduction and internal rotation.
c
From maximum shoulder abduction and internal rotation to maximum shoulder external rotation.
d
From maximum shoulder external rotation to ball release.
e
From ball release to maximum shoulder horizontal adduction.
f
Mean activity throughout the four defined phases.
MVIC = maximum voluntary isometric contraction.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Escamilla & Andrews
578
phase (tables I and II). The greatest activity of the pectoralis major, latissimus dorsi and subscapularis was during arm cocking and arm acceleration in baseball pitching (table I), while peak activity occurred in these muscles during arm acceleration and arm deceleration in football throwing (table II). The pectoralis major, latissimus dorsi and subscapularis are powerful internal rotators. These muscles contract eccentrically and help generate a shoulder internal rotation torque of »50 N m during arm cocking to slow down the externally rotating shoulder, and they contract concentrically during arm acceleration to help generate a peak shoulder internal rotation angular velocity of approximately 5000/sec.[19] The pectoralis major and subscapularis also help horizontally adduct the shoulder during arm cocking and arm acceleration, but in a different kinematic pattern compared with the baseball pitch. In football passing, the quarterback tends to ‘lead with the elbow’ as the elbow moves anterior to the trunk in achieving approximately 30 of horizontal adduction during arm cocking and arm acceleration, generating a peak horizontal adduction torque of »75 N m.[19] In contrast, in the baseball pitch the elbow remains slightly in the back of the trunk during arm cocking (»15) and slightly in front of the trunk (»5) during arm acceleration.[19] The greatest activity in the rotator cuff muscles and latissimus dorsi occurred during the arm deceleration/follow-through phase of the football throw. These muscles work to generate a peak shoulder compressive force »80% bodyweight during arm deceleration/follow-through to resist shoulder distraction, which is 20–25% less than the shoulder compressive force that is generated during baseball pitching during this phase.[19] The latissimus dorsi, posterior deltoid and infraspinatus also contract eccentrically to slow down the rapid horizontal adducting arm. Fleisig and co-authors[19] reported a shoulder horizontal abduction torque »80 N m, which is needed to help control the rate of horizontal adduction that occurs during arm deceleration/follow-through. Moreover, the peak activity that occurred in the latissimus dorsi, posterior deltoid and infraspinatus during arm deceleration/follow-through
ª 2009 Adis Data Information BV. All rights reserved.
helps resist anterior translation of the humeral head within the glenoid by, in part, generating a peak shoulder posterior force »240 N.[19] The aforementioned kinematic and kinetic differences between football passing and baseball pitching help explain the differences in muscle activity between these two activities, and they occur in part because a football weighs three times more than a baseball. Therefore, a football cannot be thrown with the same shoulder range of motion and movement speeds compared with throwing a baseball. This results in smaller loads (i.e. less shoulder forces and torques) overall applied to the shoulder in football passing compared with baseball pitching,[19] which may in part account for the greater number of shoulder injuries in baseball pitching compared with football passing. 3. Shoulder EMG during Windmill Softball Pitching Maffet et al.[4] conducted the only known study that quantified shoulder muscle firing patterns during the softball pitch. These authors used ten female collegiate softball pitchers who all threw the ‘fast pitch’ and quantified activity in the anterior and posterior deltoid, supraspinatus, infraspinatus, teres minor, subscapularis, pectoralis major and serratus anterior. The ‘fastpitch’ motion starts with the throwing shoulder extended and then as the pitcher strides forward the arm fully flexes, abducts and externally rotates and then continues in a circular (windmill) motion all the way around until the ball is released near 0 shoulder flexion and adduction. The six phases that define the pitch[4] are as follows: (i) wind-up, from first ball motion to 6 o’clock position (shoulder flexed and abducted approximately 0); (ii) from 6 o’clock to 3 o’clock position (shoulder flexed approximately 90); (iii) from 3 o’clock to 12 o’clock position (shoulder flexed and abducted approximately 180); (iv) from 12 o’clock to 9 o’clock position (shoulder abducted approximately 90); (v) from 9 o’clock position to ball release; and (vi) from ball release to completion of the pitch. The total circumduction of the arm about the shoulder from the wind-up to the follow-through Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
is approximately 450–500.[30] Moreover, this circumduction occurs while holding a 6.25–7 oz (177–198 g) ball with the elbow near full extension, which accentuates the ‘centrifugal’ distractive force acting at the shoulder. EMG results by muscle and phase during the softball pitch are shown in table III. Muscle activity was generally lowest during the wind-up and increased during the 6–3 o’clock phase as the arm began accelerating upwards. Both the supraspinatus and infraspinatus generated their highest activity during this phase. During the 6–3 o’clock phase the arm accelerates in a circular motion and achieves a peak shoulder flexion angular velocity of approximately 5000/sec.[30] The anterior deltoid was moderately active to help generate this rapid shoulder flexion angular velocity, and the serratus anterior was moderately active in helping to upwardly rotate and protract the scapula. The arm rapidly rotating upwards in a circular pattern results in a distractive force of »20–40% bodyweight, which is resisted in part by the shoulder compressive action of the supraspinatus and infraspinatus. As the arm continues its upward acceleration during the 3–12 o’clock phase, the posterior deltoids, teres minor and infraspinatus all reach their peak activity. These muscles not only help externally rotate the shoulder during this phase but also help resist the progressively increasing shoulder distractive forces, which are »50% bodyweight during this phase.[30] These muscles
579
are also in good position to resist shoulder lateral forces, which peak during this phase.[30] The arm begins accelerating downward during the 12–9 o’clock phase. It is during this phase that the shoulder begins to rapidly internally rotate 2000–3000/sec.[30] It is not surprising that the internal rotators (subscapularis and pectoralis major) exhibit high activity during this phase. High activity from the pectoralis major also helps adduct the shoulder. The subscapularis helps stabilize the humeral head and may help unload anterior capsule stress caused by the overhead and backward position of the arm as it begins accelerating forward. The serratus anterior exhibited a marked increase in activity to help stabilize the scapula and properly position the glenoid with the rapidly moving humerus. The subscapularis, pectoralis major and serratus anterior collectively generated their highest activity during the 9 o’clock to ball release phase. The serratus anterior continues to work to stabilize the scapula and properly position it in relation to the rapidly moving humerus. High subscapularis and pectoralis major activity is needed during this phase to resist distraction at the shoulder, which peaks during this phase with a magnitude of approximately bodyweight.[30,31] These muscles also help generate a peak shoulder internal rotation of approximately 4600/sec[30] and help adduct and flex the arm until the arm contacts the lateral thigh. However, not all softball pitchers exhibit the same pattern of motion
Table III. Shoulder activity by muscle and phase during the windmill softball pitcha (adapted from Maffet et al.,[4] with permission) Muscles
No. of subjects
Phase wind-up (% MVIC)
6–3 o’clock position (% MVIC)
3–12 o’clock position (% MVIC)
12–9 o’clock position (% MVIC)
10 o’clock to ball release (% MVIC)
follow-through (% MVIC)
Anterior deltoid
10
25 – 11
38 – 29
17 – 23
22 – 24
43 – 38
28 – 21
Supraspinatus
10
34 – 17
78 – 36
43 – 32
22 – 19
37 – 27
19 – 12
Infraspinatus
10
24 – 13
93 – 52
92 – 38
35 – 22
29 – 17
30 – 15
Posterior deltoid
10
10 – 5
37 – 27
102 – 42
52 – 25
62 – 29
34 – 29
Teres minor
10
8–7
24 – 25
87 – 21
57 – 21
41 – 23
44 – 11
Pectoralis major
10
18 – 11
17 – 12
24 – 18
63 – 23
76 – 24
33 – 20
Subscapularis
10
17 – 4
34 – 23
41 – 33
81 – 52
75 – 36
26 – 22
Serratus anterior
10
23 – 9
38 – 19
19 – 9
45 – 39
61 – 19
40 – 14
a
Data are given as means and standard deviations, and expressed for each muscle as a percentage of an MVIC.
MVIC = maximum voluntary isometric contraction.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Escamilla & Andrews
580
during this phase, as none of the 53 youth softball pitchers studies by Werner et al.[31] adopted the release strategy of contacting the lateral thigh at ball release. This may partially explain why the collegiate pitchers in the Maffet et al.[4] study generated relatively low posterior cuff activity and relatively low activity in general during the follow-through. With contact of the arm with the lateral thigh near ball release, the deceleration forces and torques generated by muscles to slow down the arm are much less compared with no contact of the arm with the lateral thigh. With no arm contact with the lateral thigh, shoulder compressive and related forces and torques may be higher during follow-through, as relatively high shoulder forces and torques have been reported.[30,31] However, these forces and torques are less during follow-through compared with the 9 o’clock to ball release acceleration phase. This is one major difference between overhand throwing and the ‘windmill’ type motion. In overhead throwing the deceleration phase after ball release generates greater shoulder forces and torques compared with the acceleration phase up to ball release. In softball pitching the greatest forces and torques occur during the acceleration phase of the delivery. The rapid shoulder movements and high shoulder forces that are generated during the ‘windmill fast pitch’ makes the shoulder susceptible to injury. There is also a higher risk of subacromial impingement due to the extreme shoulder flexion and abduction that occurs during the pitch. A significant number of shoulder injuries have been reported in softball pitchers, including bicipital and rotator cuff tendonitis, strain and impingement.[32] 4. Shoulder EMG during the Volleyball Serve and Spike Both the volleyball serve and spike involve an overhead throwing motion that is similar to baseball pitching and football throwing. Unlike baseball pitching and football passing, there are no known studies that have quantified the shoulder forces and torques that are generated during the volleyball serve and spike. Nevertheless, because the motion is overhead and exª 2009 Adis Data Information BV. All rights reserved.
tremely rapid, similar to baseball pitching, it is hypothesized that high shoulder forces and torques are generated, especially during the volleyball spike. To support this hypothesis, numerous injuries occur each year in volleyball, primarily involving muscle, tendon and ligament injuries during blocking and spiking.[33] It has been reported that approximately one-quarter of all volleyball injuries involve the shoulder.[33-36] Moreover, in athletes who engage in vigorous upper arm activities, shoulder pain ranks highest in volleyball players, which is largely due to the repetitive nature of the hitting motion.[33-36] Therefore, understanding muscle firing patterns of the shoulder complex is helpful in developing muscle-specific treatment and training protocols, which may both minimize injury and enhance performance. There are no known studies that have quantified muscle activity from the scapular muscles during the volley serve or spike. This is surprising given the importance of the scapular muscles in maintaining proper position of the scapula relative to the humerus. Volleyball players with shoulder pain often have muscle imbalances of the scapula muscles.[37] Therefore, the firing pattern of the scapular muscles during the volleyball serve and spike should be the focus of future research studies. Rokito et al.[6] conducted the only known study that quantified muscle firing patterns of glenohumeral muscles during the volleyball serve and spike. These authors studied 15 female college and professional volleyball players who performed both the volleyball serve and spike. The shoulder muscles quantified included the anterior deltoid, supraspinatus, infraspinatus, teres minor, subscapularis, teres major, latissimus dorsi and pectoralis major. The serve and spike motions were divided into five phases, which collectively are 1.95 sec in duration for the serve[6] and 1.11 sec for the spike:[6] (i) wind-up (comprises 39% of total serve time and 33% of total spike time) begins with shoulder abducted and extended and ends with the initiation of shoulder external rotation; (ii) cocking (comprises 20% of total serve time and 23% of total spike time) – initiation of shoulder external rotation to maximum Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
581
shoulder external rotation; (iii) acceleration (comprises 6% of total serve time and 8% of total spike time) – maximum shoulder external rotation to ball impact; (iv) deceleration (comprises 8% of total serve time and 9% of total spike time) – ball impact to when upper arm is perpendicular to trunk; and (v) follow-through (comprises 28% of total serve time and 27% of total spike time) – upper arm perpendicular to trunk to end of arm motion. Shoulder EMG results by muscle and phase during the volleyball serve and spike are shown in table IV. Similar to other overhead throwing activities, muscle activity during the serve was relatively low during the wind-up and followthrough phases. However, during the wind-up
phase of the spike, peak activity was recorded in the anterior deltoid, infraspinatus and supraspinatus. These muscles are important to help rapidly elevate the arm overhead (anterior deltoid and supraspinatus) and initiate external rotation (infraspinatus). The rotator cuff muscles are also active to help stabilize the humeral head in the glenoid fossa. During the cocking phase the shoulder rapidly externally rotates, which helps explain the high activity in the infraspinatus and teres minor during both the serve and spike. As mentioned during the section on baseball pitching, these muscles also produce a posterior force on the humerus that may help unload the anterior capsule due to the humeral head attempting to translate
Table IV. Shoulder activity by muscle and phase during the volleyball serve and spikea (adapted from Rokito et al.,[6] with permission) Muscles
Anterior deltoid
No. of subjects
Phase wind-up (% MVIC)
cocking (% MVIC)
acceleration (% MVIC)
deceleration (% MVIC)
follow-through (% MVIC)
15
Serve
21 – 11
31 – 13
27 – 22
42 – 17
16 – 16
Spike
58 – 26
49 – 19
23 – 17
27 – 10
15 – 7
Supraspinatus
15
Serve
25 – 10
32 – 18
37 – 25
45 – 13
24 – 16
Spike
71 – 31
40 – 17
21 – 27
37 – 23
27 – 15
Infraspinatus
15
Serve
17 – 10
36 – 16
32 – 22
39 – 21
13 – 11
Spike
60 – 17
49 – 16
27 – 18
38 – 19
22 – 11
Serve
7–8
44 – 20
54 – 26
30 – 23
8–9
Spike
39 – 20
51 – 17
51 – 24
34 – 13
17 – 7
Teres minor
Subscapularis
15
15
Serve
8–8
27 – 25
56 – 18
27 – 15
13 – 11
Spike
46 – 16
38 – 21
65 – 25
23 – 11
16 – 15
Teres major
15
Serve
1–1
11 – 7
47 – 24
7–8
3–3
Spike
28 – 14
20 – 11
65 – 31
21 – 18
15 – 16
Latissimus dorsi
15
Serve
1–2
9 – 18
37 – 39
6–9
3–3
Spike
20 – 13
16 – 17
59 – 28
20 – 21
15 – 10
Serve
3–6
31 – 14
36 – 14
7 – 11
7–6
Spike
35 – 17
46 – 17
59 – 24
20 – 16
21 – 12
Pectoralis major
a
15
Data are given as means and standard deviations, and expressed for each muscle as a percentage of an MVIC.
MVIC = maximum voluntary isometric contraction.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
582
anteriorly as the shoulder externally rotates. Also, the rotator cuff muscles have high activity to generate glenohumeral compression and resist distraction. The relatively high activity from the subscapularis and pectoralis major (both internal rotators) help provide support to the anterior shoulder (without such support anterior instability may ensue), as these muscles also contract eccentrically to slow down and control the rate of the rapid shoulder external rotation. An important distinction between the serve and spike occurs during the acceleration phase. During the serve the objective is not to impart maximum velocity to the ball but rather hit the ball so it ‘floats’ over the net with a parabolic trajectory in an area that would be most difficult for the opponent to return. In contrast, during the spike the primary objective is to hit the ball as hard as possible so as to convey maximum velocity to the ball. Consequently, muscle activity was higher in the powerful acceleratory muscles during the spike compared with during the serve. Because overhead throwing motions such as baseball pitching, football passing and the tennis serve achieve shoulder internal rotation angular velocities between 4000 and 7000/sec,[9,19,38] it is reasonable to assume that similar internal rotation angular velocities occur during the volleyball spike. The shoulder internal rotators (teres major, subscapularis, pectoralis major and latissimus dorsi) all generated their highest activity for both the serve and the spike in order to both internally rotate the shoulder and accelerate the arm forward. During the acceleration phase, teres minor activity peaked to provide a stabilizing posterior restraint to anterior translation. In contrast, infraspinatus activity was relatively low. The differing amounts of EMG activity between the teres minor and infraspinatus throughout the different phases of the serve and spike is interesting, especially since both the teres minor and infraspinatus provide similar glenohumeral functions and they are both located adjacent to each other anatomically. However, the spatial orientations of these two muscles are different, with the teres minor in a better mechanical position to extend the shoulder in a sagittal plane and ª 2009 Adis Data Information BV. All rights reserved.
Escamilla & Andrews
the infraspinatus in a better mechanical position to extend the shoulder in a transverse plane. There are also clinical differences between these two muscles, as they are typically not injured together but rather an isolated injury occurs to either the teres minor or infraspinatus.[2,6] These different clinical observations between the teres minor and infraspinatus are consistent with the different muscle firing patterns that occur within any given phase of overhead throwing, such as baseball pitching (table I).[2] During the deceleration phase, infraspinatus and supraspinatus activity was greatest during the serve, but not during the spike. In fact, rotator cuff activity was generally lower in the spike compared with the serve, which may be counterintuitive. For example, because a primary function of the rotator cuff is to generate shoulder compressive force to resist shoulder distraction, and since shoulder compressive forces from similar overhead throwing motions (such as baseball pitching and football passing) generate large shoulder compressive forces during this phase,[9,19] it is plausible to assume large compressive forces are also needed during the spike. The relatively low activity from the rotator cuff muscles during the spike is a different pattern compared with the moderate to high rotator cuff activity generated during the baseball pitch and football pass (tables I and II). The higher rotator cuff activity during baseball pitching and football passing is needed during this phase to resist the large distractive forces that occur at the shoulder, which are near or in excess of bodyweight. These EMG differences between varying overhead throwing motions may be due to mechanical differences between these different activities. For example, in both baseball pitching and football passing a weighted ball (5 oz [142 g] baseball and 15 oz [425 g] football) is carried in the hands throughout throwing phases but is released just prior to the beginning of the deceleration phase. With these weighted balls no longer in hand, the arm may travel faster just after ball release (beginning of deceleration phase) and thus more posterior shoulder forces and torques may be generated by the posterior musculature to slow down the rapidly moving arm. In the Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
583
volleyball spike there is no weighted implement in the hand throughout the entire motion. Moreover, when the hand contacts the ball, the ball generates an equal and opposite force on the hand, which acts to slow down the forward moving hand. Therefore, a slower moving arm may result in smaller forces and torques at the shoulder to decelerate the arm and less muscle activity. This explanation may partially explain the lower rotator cuff activity in the volleyball spike compared with baseball pitching and football passing, especially from the posterior musculature (table IV). However, a biomechanical analysis of the volleyball spike is needed to quantify shoulder forces and torques to help confirm this hypothesis. 5. Shoulder EMG during the Tennis Serve and Volley There is a scarcity of shoulder EMG data during the tennis serve and volley. Ryu and colleagues[39] conducted the only known study that extensively quantified shoulder EMG during the tennis serve. EMG data were collected during the serve from eight shoulder muscles using six male collegiate tennis players. One of the limitations of this study is there were no standard deviations reported and only a few subjects were used. The serve was divided into four phases: (i) wind-up start of service motion to ball release; (ii) cockingball release to maximum shoulder external
rotation; (iii) acceleration-maximum shoulder external rotation to racquet-ball contact; and (iv) deceleration and follow-through-racquetball contact to completion of serve. Shoulder EMG results during the serve are shown in table V. Mean EMG peaked for the infraspinatus and supraspinatus during the cocking phase. During this phase the shoulder externally rotates approximately 170 with a peak shoulder internal rotator torque of »65 N m.[38] These kinematic and kinetic data help explain the high activity from the infraspinatus, which is active to initiate shoulder external rotation during the first half of the cocking phase. The infraspinatus and supraspinatus also contract to resist shoulder distractive forces during the cocking phase. Although not quantified during the tennis serve, the shoulder compressive force needed to resist distraction is »80% bodyweight during the cocking phase in baseball pitching, which is a similar motion to the tennis serve.[11] The biceps brachii may also help generate shoulder compressive force during the cocking phase,[15] which may help explain the relatively high activity from this muscle. Pectoralis major, latissimus dorsi and subscapularis activity was greatest during the acceleration phase, as they contract to help generate a peak shoulder internal rotation angular velocity »2500/sec,[38] as well as accelerate the arm forward. Serratus anterior activity also peaked during the acceleration phase to properly
Table V. Shoulder activity by muscle and phase during the tennis servea (adapted from Ryu et al.,[39] with permission) Muscles
No. of subjects
Phase wind-up (% MVIC)
cocking (% MVIC)
acceleration (% MVIC)
deceleration and follow-through (% MVIC)
Biceps brachii
6
6
39
10
Middle deltoid
6
18
23
14
34 36
Supraspinatus
6
15
53
26
35
Infraspinatus
6
7
41
31
30
Subscapularis
6
5
25
113
63
Pectoralis major
6
5
21
115
39
Serratus anterior
6
24
70
74
53
Latissimus dorsi
6
16
32
57
48
a
Data are given as means (standard deviations not reported), and expressed for each muscle as a percentage of an MVIC.
MVIC = maximum voluntary isometric contraction.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Escamilla & Andrews
584
position the scapula relative to the rapidly moving humerus. These EMG findings during the tennis serve are similar to EMG findings during baseball pitching, which is not surprising considering there are numerous kinematic and kinetic similarities between the tennis serve and baseball pitch.[9-11,38] EMG activity during arm deceleration and follow-through demonstrated moderate to high activity, but less than the EMG observed during baseball pitching and football passing. One reason for this, as previously explained for the volleyball spike, is that the force the ball exerts against the racquet acts to slow down the arm, which may result in less posterior force and torque needed from muscle contractions. The rela-
tively high activity from the biceps brachii helps stabilize the shoulder, resist distraction and decelerate the rapid elbow extension angular velocity, which peaks at »1500/sec.[38] The moderate to high activity from the rotator cuff muscles generate compressive force to help resist shoulder distractive forces, with peak forces »75% bodyweight during the serve.[38] A few studies have examined shoulder activity during the tennis backhand and forehand.[1,39,40] Ryu and colleagues[39] collected EMG data from eight shoulder muscles using six male collegiate tennis players. This study is weakened by the low number of subjects, no standard deviations are reported and there are no statistical analyses between the forehand and backhand volleys. The
Table VI. Shoulder activity by muscle and phase during the tennis forehand and backhand volleya (adapted from Ryu et al.,[39] with permission) Muscles
Biceps brachii
No. of subjects
Phase racquet preparation (% MVIC)
acceleration (% MVIC)
deceleration and follow-through (% MVIC)
6
Forehand
17
86
53
Backhand
11
45
41
Middle deltoid
6
Forehand
27
17
20
Backhand
22
118
48
Supraspinatus
6
Forehand
22
25
14
Backhand
10
73
41
Infraspinatus
6
Forehand
29
23
40
Backhand
7
78
48
Forehand
28
102
49
Backhand
8
29
25
Subscapularis
Pectoralis major
6
6
Forehand
10
85
30
Backhand
15
29
14
Serratus anterior
6
Forehand
14
76
60
Backhand
12
45
31
Forehand
6
24
23
Backhand
4
45
10
Latissimus dorsi
a
6
Data are given as means (standard deviations not reported), and expressed for each muscle as a percentage of an MVIC.
MVIC = maximum voluntary isometric contraction.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
forehand and backhand volleys have been divided into three phases:[39] (i) racquet preparation – shoulder turn to initiation of weight transfer to front foot; (ii) acceleration-initiation of weight transfer to front foot to racquet-ball contact; and (iii) deceleration and follow-through-racquet-ball contact to completion of stroke. Shoulder EMG results from this study are shown in table VI. Muscle activity was relatively low during the racquet preparation phase, which is consistent with forehand and backhand shoulder EMG data from Chow et al.[1] Relatively large differences in muscle activity have been reported between the forehand and backhand during the acceleration phase.[1,39] High activity has been reported in the biceps brachii, anterior deltoid, pectoralis major and subscapularis during the forehand volley, but these same muscles exhibited low activity during the backhand volley.[1,39,40] The high activity during the forehand volley from the pectoralis major, anterior deltoid and subscapularis is not surprising given their role as horizontal flexors and internal rotators. However, the high activity from the biceps brachii is somewhat surprising. Morris et al.[41] also reported high biceps activity during the forehand in the acceleration phase. The biceps are in a mechanically advantageous position to horizontally flex the shoulder during the forehand motion, and they also may work to stabilize both the shoulder and elbow. Moreover, they may also help cause the slight amount of elbow flexion that occurs, or at least stabilize the elbow and keep it from extending (due to inertial forces and torques the arm applies to the forearm at the elbow as the arm rapidly horizontally flexes). The serratus anterior is also more active during the forehand compared with the backhand to help protract the scapula during the acceleration phase and help properly position the scapula relative to the rapidly moving humerus. Posterior deltoids, middle deltoids, supraspinatus, infraspinatus, latissimus dorsi and triceps brachii exhibit high activity during the backhand volley, but relatively low activity during the forehand volley.[1,39] These muscles all work synergistically during the backhand to horizontally extend and externally rotate the ª 2009 Adis Data Information BV. All rights reserved.
585
shoulder. The triceps are also active to extend the elbow and help stabilize both the shoulder and elbow. The high activity from the supraspinatus and infraspinatus help provide shoulder compressive forces to resist shoulder distraction. The supraspinatus and deltoids also help maintain the shoulder in abduction. 6. Shoulder EMG during Baseball Batting There is only one known study that has quantified muscle activity of the shoulder during baseball hitting.[7] Using the swings of 18 professional male baseball players during batting practice, these investigators quantified posterior deltoid, triceps brachii, supraspinatus and serratus anterior activity during the following swing phases: (i) wind-up – lead heel off to lead forefoot contract; (ii) pre-swing – lead forefoot contact to beginning of swing; (iii) early swing – beginning of swing to when bat was perpendicular to ground; (iv) middle swing – when bat was perpendicular to ground to when bat was parallel with ground; (v) late swing – when bat was parallel with ground to bat-ball contact; and (vi) follow-throughbat-ball contact to maximum abduction and external rotation of lead shoulder. Muscle activity was relatively low during the wind-up and follow-through phases, with EMG magnitudes generally <25% MVIC. The posterior deltoid peaked at 101% MVIC during pre-swing and then progressively decreased throughout early swing (88% MVIC), middle swing (82% MVIC) and late swing (76% MVIC). Triceps brachii activity was 46% MVIC during pre-swing, peaked at 92% MVIC during early swing, and then progressively decreased to 73% MVIC during middle swing and 38% MVIC during late swing. Both the supraspinatus and serratus anterior generated relatively moderate and constant activity from pre-swing to late swing in the range 28–39% MVIC throughout these four phases. Compared with overhand throwing, EMG data for hitting are relatively sparse, and thus it is hard to make definite conclusions. There are EMG data for only a few shoulder muscles with which to compare. Nevertheless, it does appear that both Sports Med 2009; 39 (7)
Escamilla & Andrews
586
glenohumeral and scapular muscles generate high activity during the swing, as both concentric and eccentric muscle actions are needed throughout the swing. To make it even more difficult to develop summaries of muscle firing patterns in hitting, there are currently no shoulder kinetic data in the hitting literature. The focus of future hitting studies should be on quantifying shoulder forces and torques throughout the swing, and shoulder EMG data for additional shoulder muscles, such as the infraspinatus, teres minor, pectoralis major, latissimus dorsi, biceps brachii and trapezius. 7. Shoulder EMG during the Golf Swing Several studies have examined shoulder muscle activity during the golf swing.[5,42-45] Jobe
et al.[43,44] and Pink et al.[5] used male and female professional golfers to study shoulder activity. These authors quantified both shoulder[43,44] and scapular[45] muscles of both the lead arm (left arm for a right-handed golfer) and trail arm (right arm for a right-handed golfer) and also reported no significant differences during the swing in shoulder EMG between male and female professional golfers.[44] The golf swing has been divided into five different phases:[43-45] (i) take-away – from ball address to the end of backswing; (ii) forward swing – end of backswing to when club is horizontal; (iii) acceleration – when club is horizontal to club-ball impact; (iv) deceleration – club-ball impact to when club is horizontal; and (v) follow-through – when club is horizontal to end of motion.
Table VII. Shoulder activity by muscle and phase during the golf swinga (adapted from Pink et al.,[5] with permission) Muscles
Supraspinatus
No. of subjects
Phase take-away (% MVIC)
forward swing (% MVIC)
acceleration (% MVIC)
deceleration (% MVIC)
follow-through (% MVIC)
13
Trail arm
25 – 20
14 – 14
12 – 14
7–5
7–5
Lead arm
21 – 12
21 – 15
18 – 11
28 – 20
28 – 14
Trail arm
27 – 24
13 – 16
7–8
12 – 13
9 – 10
Lead arm
14 – 12
16 – 13
27 – 25
61 – 32
40 – 24
Infraspinatus
Subscapularis
13
13
Trail arm
16 – 12
49 – 31
68 – 67
64 – 67
56 – 44
Lead arm
33 – 23
29 – 24
41 – 34
23 – 27
35 – 27
Trail arm
5–6
21 – 23
10 – 10
11 – 15
8–8
Lead arm
13 – 13
9–9
10 – 10
21 – 25
28 – 30
Anterior deltoid
Middle deltoid
13
13
Trail arm
3–3
2–3
2–5
8 – 10
8–8
Lead arm
3–3
4–6
2–2
7–8
5–3
9 – 13
17 – 16
11 – 12
9–9
8 – 14
Posterior deltoid
13
Trail arm
17 – 25
10 – 15
Lead arm
5–8
24 – 20
Trail arm
9–7
50 – 38
47 – 44
39 – 39
28 – 19
Lead arm
17 – 13
48 – 25
31 – 28
32 – 33
18 – 15
Trail arm
12 – 9
64 – 30
83 – 55
74 – 55
37 – 35
Lead arm
21 – 32
18 – 14
83 – 75
74 – 74
38 – 23
Latissimus dorsi
Pectoralis major
a
11 – 9
13
13
Data are given as means and standard deviations, and expressed for each muscle as a percentage of an MVIC.
MVIC = maximum voluntary isometric contraction.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
587
Table VIII. Scapular activity by muscle and phase during the golf swinga (adapted from Kao et al.,[45] with permission) Muscles
Phase take-away (% MVIC)
forward swing (% MVIC)
acceleration (% MVIC)
deceleration (% MVIC)
Trail arm
29 – 19
38 – 39
34 – 41
12 – 12
4–4
Lead arm
5–3
42 – 20
62 – 46
39 – 26
29 – 24
Trail arm
30 – 18
46 – 27
32 – 24
21 – 12
5–4
Lead arm
7 – 13
68 – 27
57 – 46
26 – 26
30 – 33
Levator scapulae
Rhomboids
Upper trapezius
No. of subjects
follow-through (% MVIC)
15
15
15
Trail arm
24 – 14
4–4
13 – 20
23 – 19
5–6
Lead arm
5–4
29 – 26
42 – 50
34 – 29
27 – 18
Middle trapezius
15
Trail arm
37 – 12
18 – 24
19 – 26
26 – 21
12 – 15
Lead arm
3–3
51 – 26
36 – 21
21 – 18
28 – 20
Trail arm
52 – 28
17 – 12
16 – 28
22 – 22
10 – 15
Lead arm
7 – 10
49 – 27
37 – 28
20 – 16
35 – 18
Trail arm
6–4
58 – 39
69 – 29
52 – 18
40 – 14
Lead arm
30 – 15
20 – 29
31 – 31
31 – 18
21 – 13
Trail arm
9–5
29 – 17
51 – 33
47 – 25
40 – 18
Lead arm
27 – 11
20 – 21
21 – 24
29 – 20
29 – 21
Lower trapezius
Upper serratus anterior
Lower serratus anterior
a
15
15
15
Data are given as means and standard deviations, and expressed for each muscle as a percentage of an MVIC.
MVIC = maximum voluntary isometric contraction.
Shoulder muscle activity during the golf swing is shown in table VII[5] and scapular muscle activity is shown in table VIII.[45] During the takeaway phase, muscle activity was relatively low to moderate, suggesting that lifting the arms and club up during the backswing is not a strenuous activity. The levator scapulae and lower/middle trapezius of the trail arm exhibit moderate activity during this phase to elevate and upwardly rotate the scapula, while moderate activity from the serratus anterior of the lead arm helps protract and upwardly rotate the scapula. Upper, lower and middle trapezius activities were highest during this phase compared with the other four phases. Interestingly, infraspinatus and supraspinatus activities of the trail arm were also highest during this phase but only firing »25% MVIC, which implies relatively low activity from these rotator cuff muscles throughout the golf ª 2009 Adis Data Information BV. All rights reserved.
swing. This is surprising in part because most shoulder injuries are overuse injuries that typically involve the supraspinatus or infraspinatus.[46-49] However, these rotator cuff EMG data are only for the trail arm, which may exhibit less overall rotator cuff activity throughout the swing compared with the lead arm. These data imply that rotator cuff injury risk may be higher in the lead arm, but this conclusion may not be valid because it only takes relative muscle activity into account and not other factors (such as impingement risk between shoulders). Another interesting finding is that anterior, middle and posterior deltoid activities were all relatively low throughout all phases, implying that these muscles are not used much throughout the swing. During the forward swing phase, muscle activity was also relatively low to moderate, except for relatively high activity from the subSports Med 2009; 39 (7)
Escamilla & Andrews
588
scapularis, pectoralis major, latissimus dorsi and serratus anterior of the trail arm to adduct and internally rotate the trail arm and protract the scapula. There was also relatively high activity from the rhomboids and middle/lower trapezius of the lead arm to help retract and stabilize the scapula. Muscle activity during the acceleration phase was higher overall compared with the forward swing phase. The subscapularis, pectoralis major, latissimus dorsi and serratus anterior of the trail arm demonstrated high activity during the acceleration phase to continue adducting and internally rotating the trail arm. These muscles may be the most important ‘power’ muscles of the upper extremity to help accelerate the arm during the acceleration phase of the downswing. In addition, using a short or long backswing may affect shoulder activity during the acceleration phase. Slightly greater pectoralis major and latissimus dorsi activity has been reported during the acceleration phase when a short backswing was used compared with a long backswing, pointing to the conclusion that shoulder injury risk may increase over time.[42] During the deceleration phase the subscapularis, pectoralis major, latissimus dorsi and serratus anterior of the trail arm continued to demonstrated high activity, although now the muscle action was more eccentric and slightly smaller in magnitude compared with the acceleration phase. Low to moderate activity occurred from the scapular muscles of the lead arm, while high pectoralis major and infraspinatus activity occurred in the lead arm. Muscle activity generally decreased from the deceleration phase to the follow-through phase. 8. Conclusions This review reports shoulder muscle activity, and when available shoulder kinematics and kinetics, during a variety of upper extremity sports. During overhead throwing, high rotator cuff muscle activity was generated to help resist the high shoulder distractive forces of »80–120% bodyweight during the arm cocking and deceleration phases. During arm cocking, peak rotaª 2009 Adis Data Information BV. All rights reserved.
tor cuff activity is 49–99% MVIC in baseball pitching and 41–67% MVIC in football throwing. During arm deceleration, peak rotator cuff activity is 37–84% MVIC in baseball pitching and 86–95% MVIC in football throwing. Peak rotator cuff activity is also high in the windmill softball pitch (75–93% MVIC), the volleyball serve and spike (54–71% MVIC), the tennis serve and volley (40–113% MVIC), baseball hitting (28–39% MVIC) and the golf swing (28–68% MVIC). Peak scapular muscle activity is also high during the arm cocking and arm deceleration phases of baseball pitching, with peak serratus anterior activity 69–106% MVIC, peak upper, middle and lower trapezius activity 51–78% MVIC, peak rhomboids activity 41–45% MVIC and peak levator scapulae activity 33–72% MVIC. Moreover, peak serratus anterior activity was »60% MVIC during the windmill softball pitch, »75% MVIC during the tennis serve and forehand and backhand volley, »30–40% MVIC during baseball hitting, and »70% MVIC during the golf swing. In addition, during the golf swing, peak upper, middle and lower trapezius activity was 42–52% MVIC, peak rhomboids activity was »60% MVIC, and peak levator scapulae activity was »60% MVIC. Understanding when and how much the shoulder muscles are active during upper extremity sports is helpful to physicians, therapists, trainers and coaches in providing appropriate treatment, training and rehabilitation protocols to these athletes, as well as help better understand the injury mechanism. Acknowledgements No funding was provided for the preparation of this review, and the authors have no conflicts of interest that are directly relevant to the content of this review.
References 1. Chow JW, Carlton LG, Lim YT, et al. Muscle activation during the tennis volley. Med Sci Sports Exerc 1999; 31 (6): 846-54 2. DiGiovine NM, Jobe FW, Pink M, et al. EMG of upper extremity in pitching. J Shoulder Elbow Surg 1992; 1: 15-25
Sports Med 2009; 39 (7)
Shoulder Muscle Activity in Upper Extremity Sports
3. Kelly BT, Backus SI, Warren RF, et al. Electromyographic analysis and phase definition of the overhead football throw. Am J Sports Med 2002; 30 (6): 837-44 4. Maffet MW, Jobe FW, Pink MM, et al. Shoulder muscle firing patterns during the windmill softball pitch. Am J Sports Med 1997; 25 (3): 369-74 5. Pink M, Jobe FW, Perry J. Electromyographic analysis of the shoulder during the golf swing. Am J Sports Med 1990; 18 (2): 137-40 6. Rokito AS, Jobe FW, Pink MM, et al. Electromyographic analysis of shoulder function during the volleyball serve and spike. J Shoulder Elbow Surg 1998; 7 (3): 256-63 7. Shaffer B, Jobe FW, Pink M, et al. Baseball batting: an electromyographic study. Clin Orthop Relat Res 1993; 292 (292): 285-93 8. Lawrence JH, De Luca CJ. Myoelectric signal versus force relationship in different human muscles. J Appl Physiol 1983; 54 (6): 1653-9 9. Escamilla RF, Barrentine SW, Fleisig GS, et al. Pitching biomechanics as a pitcher approaches muscular fatigue during a simulated baseball game. Am J Sports Med 2007; 35 (1): 23-33 10. Escamilla RF, Fleisig GS, Barrentine SW, et al. Kinematic comparisons of throwing different types of baseball pitches. J Appl Biomech 1998; 14 (1): 1-23 11. Fleisig GS, Andrews JR, Dillman CJ, et al. Kinetics of baseball pitching with implications about injury mechanisms. Am J Sports Med 1995; 23 (2): 233-9 12. Fleisig GS, Escamilla RF, Andrews JR, et al. Kinematic and kinetic comparison between baseball pitching and football passing. J Appl Biomech 1996; 12: 207-24 13. Jobe FW, Kvitne RS, Giangarra CE. Shoulder pain in the overhand or throwing athlete: the relationship of anterior instability and rotator cuff impingement. Orthop Rev 1989; 18 (9): 963-75 14. Meister K. Injuries to the shoulder in the throwing athlete: part one, biomechanics/pathophysiology/classification of injury. Am J Sports Med 2000; 28 (2): 265-75 15. Glousman R, Jobe F, Tibone J, et al. Dynamic electromyographic analysis of the throwing shoulder with glenohumeral instability. J Bone Joint Surg [Am] 1988; 70 (2): 220-6 16. Gowan ID, Jobe FW, Tibone JE, et al. A comparative electromyographic analysis of the shoulder during pitching: professional versus amateur pitchers. Am J Sports Med 1987; 15 (6): 586-90 17. Jobe FW, Moynes DR, Tibone JE, et al. An EMG analysis of the shoulder in pitching: a second report. Am J Sports Med 1984; 12 (3): 218-20 18. Jobe CM. Special properties of living tissue that affect the shoulder in athletes. Clin Sports Med 1983; 2 (2): 271-80 19. Fleisig GS, Barrentine SW, Escamilla RF, et al. Biomechanics of overhand throwing with implications for injuries. Sports Med 1996; 21 (6): 421-37 20. Toyoshima S, Hoshikawa T, Miyashita M, et al. Contribution of the body parts to throwing performance. In: Nelson RC, Morehouse CA, editors. Biomechanics IV. Baltimore (MD): University Part Press, 1974: 169-74
ª 2009 Adis Data Information BV. All rights reserved.
589
21. Jobe FW, Tibone JE, Perry J, et al. An EMG analysis of the shoulder in throwing and pitching: a preliminary report. Am J Sports Med 1983; 11 (1): 3-5 22. Escamilla R, Fleisig G, Barrentine S, et al. Kinematic and kinetic comparisons between American and Korean professional baseball pitchers. Sports Biomech 2002; 1 (2): 213-28 23. Werner SL, Fleisig GS, Dillman CJ, et al. Biomechanics of the elbow during baseball pitching. J Orthop Sports Phys Ther 1993; 17 (6): 274-8 24. Feltner M, Dapena J. Dynamics of the shoulder and elbow joints of the throwing arm during a baseball pitch. Inter J Sport Biomech 1986; 2: 235-59 25. Rodosky MW, Harner CD, Fu FH. The role of the long head of the biceps muscle and superior glenoid labrum in anterior stability of the shoulder. Am J Sports Med 1994; 22 (1): 121-30 26. De Wilde L, Plasschaert F, Berghs B, et al. Quantified measurement of subacromial impingement. J Shoulder Elbow Surg 2003; 12 (4): 346-9 27. Pappas AM, Zawacki RM, Sullivan TJ. Biomechanics of baseball pitching: a preliminary report. Am J Sports Med 1985; 13 (4): 216-22 28. Solem-Bertoft E, Thuomas KA, Westerberg CE. The influence of scapular retraction and protraction on the width of the subacromial space: an MRI study. Clin Orthop Relat Res 1993; 296: 99-103 29. Roberts EM. Cinematography in biomechanical investigation. Proceedings of the C.I.C. Symposium on Biomechanics. Chicago (IL): The Athletic Institute, 1971 30. Barrentine SW, Fleisig GS, Whiteside JA, et al. Biomechanics of windmill softball pitching with implications about injury mechanisms at the shoulder and elbow. J Orthop Sports Phys Ther 1998; 28 (6): 405-15 31. Werner SL, Guido JA, McNeice RP, et al. Biomechanics of youth windmill softball pitching. Am J Sports Med 2005; 33 (4): 552-60 32. Loosli AR, Requa RK, Garrick JG, et al. Injuries to pitchers in women’s collegiate fast-pitch softball. Am J Sports Med 1992; 20 (1): 35-7 33. Watkins J, Green BN. Volleyball injuries: a survey of injuries of Scottish National League male players. Br J Sports Med 1992; 26 (2): 135-7 34. Chandler TJ, Kibler WB, Uhl TL, et al. Flexibility comparisons of junior elite tennis players to other athletes. Am J Sports Med 1990; 18 (2): 134-6 35. Schafle MD. Common injuries in volleyball: treatment, prevention and rehabilitation. Sports Med 1993; 16 (2): 126-9 36. Schafle MD, Requa RK, Patton WL, et al. Injuries in the 1987 national amateur volleyball tournament. Am J Sports Med 1990; 18 (6): 624-31 37. Kugler A, Kruger-Franke M, Reininger S, et al. Muscular imbalance and shoulder pain in volleyball attackers. Br J Sports Med 1996; 30 (3): 256-9 38. Elliott B, Fleisig G, Nicholls R, et al. Technique effects on upper limb loading in the tennis serve. J Sci Med Sport 2003; 6 (1): 76-87 39. Ryu RK, McCormick J, Jobe FW, et al. An electromyographic analysis of shoulder function in tennis players. Am J Sports Med 1988; 16 (5): 481-5
Sports Med 2009; 39 (7)
590
40. Adelsberg S. The tennis stroke: an EMG analysis of selected muscles with rackets of increasing grip size. Am J Sports Med 1986; 14 (2): 139-42 41. Morris M, Jobe FW, Perry J, et al. Electromyographic analysis of elbow function in tennis players. Am J Sports Med 1989; 17 (2): 241-7 42. Bulbulian R, Ball KA, Seaman DR. The short golf backswing: effects on performance and spinal health implications. J Manipulative Physiol Ther 2001; 24 (9): 569-75 43. Jobe FW, Moynes DR, Antonelli DJ. Rotator cuff function during a golf swing. Am J Sports Med 1986; 14 (5): 388-92 44. Jobe FW, Kvitne RS, Giangarra CE. Shoulder pain in the overhand or throwing athlete: the relationship of anterior instability and rotator cuff impingement [published erratum appears in Orthop Rev 1989 Dec; 18 (12): 1268]. Orthop Rev 1989; 18 (9): 963-75
ª 2009 Adis Data Information BV. All rights reserved.
Escamilla & Andrews
45. Kao JT, Pink M, Jobe FW, et al. Electromyographic analysis of the scapular muscles during a golf swing. Am J Sports Med 1995; 23 (1): 19-23 46. Hamilton CD, Glousman RE, Jobe FW, et al. Dynamic stability of the elbow: electromyographic analysis of the flexor pronator group and the extensor group in pitchers with valgus instability. J Shoulder Elbow Surg 1996; 5 (5): 347-54 47. Choi CH, Kim SK, Jang WC, et al. Biceps pulley impingement: Arthroscopy 2004; 20 Suppl. 2: 80-3 48. McHardy A, Pollard H, Luo K. Golf injuries: a review of the literature. Sports Med 2006; 36 (2): 171-87 49. Wiesler ER, Lumsden B. Golf injuries of the upper extremity. J Surg Orthop Adv 2005; 14 (1): 1-7
Correspondence: Prof. Rafael F. Escamilla, Department of Physical Therapy, California State University, 6000 J Street, Sacramento, CA 95819-6020, USA. E-mail:
[email protected]
Sports Med 2009; 39 (7)
Sports Med 2009; 39 (7): 591-605 0112-1642/09/0007-0591/$49.95/0
REVIEW ARTICLE
ª 2009 Adis Data Information BV. All rights reserved.
Effect of Sensorimotor Training on Morphological, Neurophysiological and Functional Characteristics of the Ankle A Critical Review Maarten D.W. Hupperets,1 Evert A.L.M. Verhagen1 and Willem van Mechelen1,2 1 Department of Public and Occupational Health, Institute for Research in Extramural Medicine, VU University Medical Centre, Amsterdam, the Netherlands 2 Body@Work Research Centre for Physical Activity, Work and Health, TNO VUmc, Amsterdam, the Netherlands
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Literature Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Study Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Quality Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Overview of Selected Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Effects of Sensorimotor Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Neurophysiological Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Morphological Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Combination of Neurophysiological and Morphological Changes . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Functional Changes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Training versus Learning Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract
591 593 593 593 594 594 594 598 599 600 602 603
Sensorimotor training is effective in preventing ankle sprain recurrences, but the pathway through which this effect occurs is unknown. Biomechanical and neurophysiological analyses of sensorimotor training leading to functional changes of the ankle are needed to establish this pathway. This article reviews the effect of sensorimotor training on morphological, neurophysiological and functional characteristics of the ankle. A MEDLINE and CINAHL computerized literature search was conducted to search for relevant articles. A study was included if (i) the study contained research questions regarding the effect of sensorimotor training on mechanical, neurophysiological, and/or functional ankle functioning; (ii) the study dealt with subjects with a history of ankle sprain; (iii) the study contained a control group; (iv) the results contained measures of mechanical, neurophysiological or functional insufficiencies as study outcome; and (v) the study met a predefined cut-off score set for methodological quality.
Hupperets et al.
592
Results on joint position sense and muscle reaction times showed a learning effect of repeated measures and not a training effect. Decrements of postural sway after sensorimotor training were mainly attributable to a learning effect as well. Effects on muscle strength were not found. Evidence for an effect of sensorimotor training on neurophysiological, morphological and functional characteristics is limited, if present at all. Thus, the pathway of sensorimotor training remains unclear. Future studies need to focus on (i) differentiating between morphological, physiological and functional changes; (ii) larger sample sizes with a priori sample size calculations; (iii) correspondence between training and test method; (iv) using measures other than postural sway more closely linked to functional stability; and (v) using a longer follow-up period than 6 weeks.
With an incidence of about 25% of all injuries across all sports, the most common injury is the ankle injury.[1-5] Of all ankle injuries, up to 95% involve the lateral ankle ligaments, i.e. acute lateral ankle sprain, causing a partial or complete rupture of the anterior talofibular ligament and in some cases also the calcaneofibular ligament.[6] It has been well documented that athletes who experience an ankle sprain have a higher risk of re-injury within 1 year post-injury.[7-10] This increased injury risk after an initial ankle sprain is generally thought to be caused by a proprioceptive impairment in the ankle due to trauma to mechanoreceptors of the ankle ligaments after an ankle sprain.[11] Partly based on this rationale, sensorimotor training is widely used for rehabilitation after an ankle sprain, and is thought to improve proprioception by re-establishing and strengthening the protective reflexes of the ankle.[12,13] As stated by Ashton Miller et al.[14] and Taube et al.,[15] multiple terms are used to summarize balance exercises aimed at improving balance. Although the term ‘proprioceptive training’ is widely used, by definition, proprioception is a purely afferent sensory modality, discarding adaptations on the motor side.[15] The exercises described in the literature, and therefore also in this review, are primarily motor (efferent) tasks. Other terms such as ‘neuromuscular training’,[16] or preferably ‘sensorimotor training’,[17,18] have been used extensively in the literature to describe the afferent and efferent aspects of these processes. In a variety of sports, multiple studies[19-29] have looked at the effectiveness of sensorimotor ª 2009 Adis Data Information BV. All rights reserved.
training for the prevention of ankle sprains. A common finding in these studies is that sensorimotor training reduces the increased injury risk for ankle sprains in athletes with a previous injury to the same level as athletes without any history of ankle sprains.[24,27] Athletes without a previous injury do not seem to benefit from sensorimotor training.[30] Thereby, these studies provide indirect evidence that sensorimotor training indeed improves ankle proprioception after an initial ankle sprain. However, a ‘true’ effect on ankle proprioception due to sensorimotor balance board training can only be established through biomechanical and neurophysiological analyses, looking at the pathway of morphological (changes in ankle form and structure) and neurophysiological changes (changes in nervous system function) of the ankle, leading to functional effects (changes in physiological activity of the ankle, e.g. postural sway). Although impaired proprioception, impaired neuromuscular control, strength deficits and impaired postural control are all denominated ‘functional insufficiencies’,[31] logically, changes in morphology and physiology cause a change in function of the ankle, which in turn will show up in a decrease in for instance postural sway. Hertel’s[31] interpretation of all the abovementioned changes as ‘functional insufficiencies’ is widely used,[27,32-39] while only postural sway can be considered a true functional measure of proprioception. In contrast to the epidemiological studies on the preventive effect of sensorimotor training, Sports Med 2009; 39 (7)
Sensorimotor Training for Ankle Injury Recovery
which are characterized by large cohorts and prospective study design with a long follow-up, most biomechanical and neurophysiological studies rely on small sample sizes and mixed study designs. Therefore, it is not surprising that the number of these studies reporting changes in ankle functioning due to a sensorimotor training programme matches the number of studies failing to show such changes.[40] Thereby, despite an abundance of epidemiological evidence on the secondary preventive effect of sensorimotor training, the biomechanical and/or neurophysiological pathway through which such a training programme affects injury risk remains unclear. For this reason, a critical review of the literature was performed to answer the following two questions: 1. What is the effect of sensorimotor training on morphological and neurophysiological characteristics in subjects with a previous ankle sprain? 2. What is the effect on measurable functional properties such as postural sway?
1. Literature Search 1.1 Study Selection
A MEDLINE and CINAHL computerized literature search was conducted for published studies relating to sensorimotor training and mechanical, neurophysiological and functional insufficiencies of the ankle. The search period ranged from January 1980 to December 2007. The following keywords were used in the search: ‘sensorimotor training’, ‘balance training’, ‘neuromuscular training’, ‘proprioceptive training’ or ‘proprioception training’, in combination with ‘ankle sprain’ or ‘ankle instability’, and ‘functional’ or ‘mechanical’. A search for relevant articles was also performed in reference lists of the identified studies. A study was included in the review if: (1) the study contained research questions regarding the effect of sensorimotor training on biomechanical, neurophysiological and/or functional ankle functioning; ª 2009 Adis Data Information BV. All rights reserved.
593
(2) the study dealt with subjects with a history of ankle sprain; (3) the study contained a control group; (4) the results contained measures of biomechanical, neurophysiological or functional insufficiencies as study outcome; and (5) the study met a predefined cut-off score set for methodological quality. Criteria 1–4 were used as a pre-selection measure, whereas criterion 5 was used as a quality assessment of eligible studies. The initial search yielded a total of 58 studies, of which 24 were relevant and of potential interest. None of these studies dealt with mechanical ankle functioning. Twelve of 19 studies met criteria 1 through 4, and were considered relevant for this review (table I). 1.2 Quality Assessment
A total of 11 predefined criteria were used in order to assess the quality of studies (table II). These criteria were adapted from Verhagen et al.[56] and customized to cover the topic of this review. The 12 relevant studies were reviewed on design and methodology independently by two reviewers (EALMV and MDWH). Each item of a selected study that met a criterion was assigned a value of ‘1’ (positive), if the item did not meet a criterion or was not described at all a ‘0’ was assigned. This made the highest possible score ‘11’. In case of a difference in opinion on items, both reviewers tried to reach consensus. If no consensus was reached, a third reviewer (WvM) was to make the final decision. This latter, however, did not occur. In order to establish the reliability and proper use of this set of predefined criteria, the inter-rater agreement, expressed as Cohen’s Kappa, was calculated. Studies scoring ‡60% (‘7’ or more) of the maximum possible score were considered to be of sufficient methodological quality and were taken into further analysis. The choice for this cut-off score is in agreement with Verhagen et al.[56] and Borghouts et al.,[57] who used a similar arbitrary choice of cut-off score and stated this to be the best way to make a discrimination between ‘highquality’ and ‘low-quality’ studies. The results of Sports Med 2009; 39 (7)
Hupperets et al.
594
Table I. Score for pre-selection criteria 1–4a Study (year)
Training
History
Control
Relevant outcome measure
Akhbari et al.[41] (2007)
1
1
0
1; postural sway; muscle reaction time
Bernier and Perrin[32] (1998)
1
1
1
1; postural sway, joint position sense
Docherty et al.[42] (2006)
0
1
1
1; postural sway
Eils and Rosenbaum[33] (2001)
1
1
1
1; postural sway, joint position sense, muscle reaction time
Gauffin et al.[43] (1988)
1
1
0
1; postural sway
Hale et al.[44] (2007)
1
1
1
1; postural sway
Hess et al.[34] (2001)
1
1
1
1; postural sway
Høiness et al.[35] (2003)
1
1
1
1; postural sway, strength
Holme et al.[27] (1999)
1
1
1
1; postural sway, joint position sense, strength
Kaminski et al.[45] (1999)
0
1
1
1; strength
Kaminski et al.[36] (2003)
1
1
1
1; strength
Kidgell et al.[37] (2007)
1
1
1
1; postural sway
Matsusaka et al.[46] (2001)
1
1
0
1; postural sway
Michell et al.[47] (2006)
1
1
0
1; postural sway
Osborne et al.[48] (2001)
1
1
0
1; muscle reaction time
Pintsaar et al.[49] (1996)
1
1
0
1; postural sway
Powers et al.[38] (2004)
1
1
1
1; postural sway
Ross and Guskiewicz[50] (2006)
1
1
1
1; postural sway
Ross et al.[51] (2007)
1
1
1
1; postural sway
Rozzi et al.[52] (1999)
1
1
0
1; postural sway
Sekir et al.[53] (2007)
0
1
1
1; postural sway, joint position sense, strength
Sheth et al.[54] (1997)
1
0
1
1; muscle reaction time
Verhagen et al.[39] (2005)
1
1
1
1; postural sway
Willems et al.[55] (2002)
0
1
1
1; joint position sense, strength
a
1 = positive score, 0 = negative score.
the methodological quality score of the individual studies are presented in table III. After both reviewers had scored the studies, there was disagreement on three (2.3%) of the 132 criteria scored. This made the inter-rater agreement high (k = 0.89). All studies met the predefined cut-off score of 60%. 1.3 Overview of Selected Studies
Characteristics of the reviewed studies are presented in table IV. The next section presents the effect of sensorimotor training on neurophysiological changes (i.e. joint position sense), morphological changes (i.e. strength), or a combination measure of both entities (i.e. muscle reaction time). Next, changes in postural sway as a measurable of functional effect of sensorimotor training are presented. ª 2009 Adis Data Information BV. All rights reserved.
Although limited on several points, all studies scored high on methodological quality. All but two studies[33,35] scored negative on homogeneity of subject characteristics. In four studies[32,36,38,50] a description of subject characteristics was missing. In other studies a significant difference in mean height,[34,37] mean age,[34] total number of positive anterior drawer signs,[27] or sex[39] was found between groups. The lack of definition of history of injury,[35,38] an insufficiently detailed description of the compliance to the programme[32,37,38,50,51] or randomization procedure[33,50] led to negative scores as well. 2. Effects of Sensorimotor Training 2.1 Neurophysiological Changes
Holme et al.[27] looked at the influence of sensorimotor training on the ability to reproduce Sports Med 2009; 39 (7)
Sensorimotor Training for Ankle Injury Recovery
595
impossible to assess whether progress was achieved through the training programme, because baseline data are missing. Baseline position sense of the ankle could well have been different between both groups, because of the significantly greater number of positive anterior drawer signs in the training group compared with the control group at baseline. Bernier and Perrin[32] found an improvement in joint position sense after 6 weeks of sensorimotor training for subjects in the training group compared with baseline. A similar improvement was found for the control group as well, which indicates a learning effect of repeated testing. No significant differences between groups were found at 6 weeks, which can have different causes: e.g. the specificity of the training compared with the testing method (the coordination and strength exercises were carried out standing up, whereas the test was performed lying down, bearing no bodyweight). A second reason could be that a training period of 6 weeks – three times 10 minutes per week – is too short to bring about physiological changes at the CNS level. Eils and Rosenbaum[33] found an improved position sense in a group of athletes with chronic ankle instability undergoing sensorimotor training, whereas the control group showed no improvement. It was not clear if significant position sense differences between groups existed at
Table II. Quality assessment: criteria list for the assessment of the methodological score of studies on effects of sensorimotor training on functional and mechanical insufficiencies Criterion (a) Is a randomization procedure mentioned? (b) Are the intervention and control groups homogenous regarding the subject characteristics? (c) Is a definition for ‘history of injury’ given? (d) Are testing procedures described and performed in sufficient detail? 1 test procedure; 2 standardized measurement (e) Are intervention procedures described and performed in sufficient detail? 1 applied training procedure; 2 time span of intervention; 3 control of compliance to the intervention (f) Are the research design and statistical analysis sufficient? 1 statistical analysis is consistent with the research design; 2 corrected for accurate variables; 3 are all relevant statistical outcomes presented (e.g. mean, SD, p-value)?
joint angles. They compared a group of subjects given general instructions on ankle sprain rehabilitation to a group given the same instruction and 4 months of sensorimotor training. Since subjects were included 5 days after injury, no postural sway baseline measurements could be determined. No improved position sense was found in the training group compared with the control group after 6 weeks and 4 months. Furthermore, no differences in position sense were found between injured and non-injured ankles. From these findings the conclusion can be drawn that sensorimotor training was not effective in improving joint position sense. However, it is
Table III. Methodological quality score of relevant studies for the criteria A to F listed in table IIa Total score
% of total
Study (year)
A
B
C
D1
D2
E1
E2
E3
F1
F2
F3
Bernier and Perrin[32] (1998)
1
0/u
1
1
1
1
1
0/u
1
1
1
9
Eils and Rosenbaum[33] (2001)
0
1
1
1
1
1
1
1
1
1
1
10
91
Hale et al.[44] (2007)
1
1
1
1
1
1
1
1
1
1
1
11
100
Hess et al.[34] (2001)
1
0
1
1
1
1
1
1
1
1
1
10
91
Høiness et al.[35] (2003)
1
1
0
1
1
1
1
1
1
1
1
10
91
Holme et al.[27] (1999)
1
0
1
1
1
1
1
1
1
1
1
10
91 91
82
Kaminski et al.[36] (2003)
1
0/u
1
1
1
1
1
1
1
1
1
10
Kidgell et al.[37] (2007)
1
0
1
1
1
1
1
0
1
1
1
9
82
Powers et al.[38] (2004)
1
0/u
0
1
1
1
1
0
1
1
1
8
73
Ross and Guskiewicz[50] (2006)
0
0/u
1
1
1
1
1
0/u
1
1
1
8
73
Ross et al.[51] (2007)
1
1
1
1
1
1
1
0/u
1
1
1
10
91
Verhagen et al.[39] (2005)
1
0
1
1
1
1
1
1
1
1
1
10
91
a
1 = positive score, 0 = negative score, u = unknown.
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Hupperets et al.
596
Table IV. Characteristics of relevant and methodologically adequate studies Study (year)
Subjects
Training
Tests
Outcome
Bernier and Perrin[32] (1998)
Functional instability: 14 control, 14 sham, 17 experimental
6 wk, 3 times a week. Sensorimotor training
Postural sway: single-leg stance on static and moving force plate Joint position sense: angle reproduction test on ankle footplate
Postural sway: main effect on sway index and equilibrium score for test in all groups (F(1,42) = 11.07, p = 0.002; resp. F(1,42) = 6.63, p = 0.014). Training group significantly improved on postural sway and balance in the stable platform condition (eyes closed; p < 0.05) and the moving platform condition (eyes open; p < 0.05) Joint position sense: no significant differences between groups at baseline and after training. All subjects performed better on the test after 6 weeks (F(1,42) = 5.46, p = 0.024)
Eils and Rosenbaum[33] (2001)
Chronic instability: 10 control, 20 experimental
6 wk, once a week. Sensorimotor training
Postural sway: single-leg stance on force plate Joint position sense: angle reproduction test on footplate Muscle reaction time: reaction after simulated ankle sprain on trap door
Postural sway: sway of centre of gravity and sway distance improved significantly in the intervention group and the control group. The intervention group improved significantly on static balance in mediolateral direction (pre mean – SD: 22.3 – 4.8 mm; post 20.4 – 2.8; p < 0.01). The control group improved significantly on static balance in anteroposterior direction (pre mean – SD: 30.3 – 7.9 mm; post 27.8 – 7.8; p < 0.05) Joint position sense: training group improved significantly after 6 weeks (pre mean error – SD: 2.0 – 0.6; post 1.5 – 0.4; p < 0.01) Muscle reaction time: significant prolongation of reaction time after 6 weeks for peroneus longus (pre mean – SD: 61.6 – 6.5 msec; post 64.8 – 6.2 msec; p < 0.001) and brevis muscles (pre mean – SD: 66.9 – 6.8 msec; post 70.4 – 6.0 msec; p < 0.001)
Hess et al.[34] (2001)
Functional instability: 10 control, 10 experimental
4 wk, 3 times a week. Agility training
Postural sway: single-leg stance on balance system
Postural sway: no effect of agility training on balance was found
Hale et al.[44] (2007)
Chronic instability: 29 subjects randomly divided over intervention, control
4 wk; variable intensity. Strength and sensorimotor training
Postural sway: single-leg stance on force plate SEBT, FADI-Sport
Postural sway: At baseline and at 4 weeks no significant postural sway differences were found SEBT: no differences between groups at baseline. At follow-up, the intervention group showed significant improvements compared with control in the posteromedial (p = 0.013), posterolateral (p = 0.011) and lateral directions (p = 0.018), as well as in the mean of all eight reach directions (p = 0.019) FADI: no differences between groups at baseline. At follow-up the intervention Continued next page
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Sensorimotor Training for Ankle Injury Recovery
597
Table IV. Contd Study (year)
Subjects
Training
Tests
Outcome group showed more improvement compared with baseline (FADI: p = 0.003; FADI-Sport: p = 0.009).
Høiness et al.[35] (2003)
Mechanical instability: 9 control, 10 experimental
6 wk, 3 times a week. High-intensity aerobic training
Postural sway: single-leg stance on balance system Strength: peak eversion torque on isokinetic dynamometer
Difference in postural sway at baseline (mean – SD: int. 72.5 – 10.7% of max; con 56.1 – 32.0, p-value data missing); No significant differences in strength at baseline Postural sway: after training both groups kept balance at a significantly higher speed of the moving platform (int.: p = 0.005, CI 2.5, 12.5; con: p = 0.018, CI 2.3, 21.0). Strength: eversion torque increased peak inversion torque significantly after bidirectional pedal training. Traditional pedal showed no increase. Difference between groups was not significant (p = 0.264, resp. p = 0.239).
Holme et al.[27] (1999)
Recent ankle sprain: 46 control, 46 experimental
4 months, 2 times a week. Balance training as part of supervised rehabilitation
Postural sway: single-leg stance on force plate Joint position sense: angle reproduction test on footplate Strength: peak eversion torque on isokinetic dynamometer
No baseline measurements were executed Postural sway: no significant differences between groups after 6 weeks and 4 months. Significant difference between affected and nonaffected side in both groups at 6 weeks (mean – SD: int. 219 – 15 cm resp. 197 – 11, p < 0.001; con 237 – 13 resp. 196 – 9, p < 0.001), not after 4 months Joint position sense: no significant difference after 6 weeks and 4 months Strength: no significant differences between both groups after 6 weeks and 4 months
Kaminski et al.[36] (2003)
Functional instability: 38 subjects randomly divided over control, strength, sensorimotor, combination
6 wk, 3 times a week. Strength and sensorimotor training
Strength: peak torques for dorsiflexion, plantar flexion, inversion and eversion on isokinetic dynamometer
Strength: no significant increase in E/I ratios (at 30/sec: F(3,34) = 1.434, p = 0.250; at 120/sec: F(3,34) = 2.123, p = 0.116) and no significant differences between the four groups at the end of the training protocol (at 30/sec: F(3,34) = 0.098, p = 0.961; at 120/sec: F(3,34) = 1.434, p = 0.929)
Kidgell et al.[37] (2007)
Functional instability: 7 control, 7 dura disc, 6 mini-trampoline
6 wk, 3 times a week. Balance training
Postural sway: single-leg stance on force plate
Postural sway: significant improvement in postural sway after 6 weeks for both training groups compared with the control group (pre mean – SD: con 36.9 – 9.9 mm; post 36.7 – 8.2, MT 56.8 – 20.5 resp. 33.3 – 8.5, p = 0.003, DT 41.3 – 2.6 resp. 27.2 – 4.8, p = 0.003)
Powers et al.[38] (2004)
Functional instability: 38 subjects randomly divided over control, strength,
6 wk, 3 times a week. Strength and sensorimotor training
Postural sway: single-leg stance on force plate
Postural sway: after 6 weeks of training there were no significant differences between groups with respect to centre of pressure excursion (group · test: Continued next page
ª 2009 Adis Data Information BV. All rights reserved.
Sports Med 2009; 39 (7)
Hupperets et al.
598
Table IV. Contd Study (year)
Subjects
Training
Tests
Outcome
sensorimotor, combination
p = 0.737. Pre mean – SD: con 9.11 – 1.14 cm; post 8.96 – 2.25, str 8.42 – 1.15 resp. 8.28 – 0.74, prop 8.82 – 4.37 resp. 7.92 – 1.80; comb 8.23 – 0.90 resp. 8.72 – 0.88) Postural sway: after 6 weeks both training groups significantly improved TTS compared with baseline. Compared with the control group there was no significant improvement
Ross and Guskiewicz[50] (2006)
Functional instability: 30 subjects assigned to conventional, stimulus, control
6 wk, 5 times a week. Coordination training with or without stochastic resonance stimulation
Postural sway: dynamic through time to stabilization
Ross et al.[51] (2007)
Functional instability: 30 subjects assigned to conventional, stimulus, control
6 wk, 5 times a week. Coordination training with or without stochastic resonance stimulation
Postural sway: single-leg stance on force plate
Postural sway: no group differences at baseline. After 6 weeks no difference in postural sway between conventional training group and control. The stimulation group had reduced post-test means compared with pooled conventional and control
Verhagen et al.[39] (2005)
Previous ankle sprain: 11 control, 11 experimental, 8 volleyball
5.5 wk, 2 times a week. Sensorimotor training
Postural sway: single-leg stance on force plate
No significant differences between groups at baseline Postural sway: no significant differences between groups in centre of pressure excursion and no differences between the groups after the training period
comb = combination training group; con = control group; DT = dura disc training group; dura disc = sensorimotor training tool; E/I = eversion to inversion; FADI = Foot and Ankle Disability Index; int = intervention group; MT = mini-trampoline group; prop = proprioceptive training group; resp. = respectively; SEBT = Star Excursion Balance Test; str = strength training group; TTS = time-to-stabilization.
baseline, but baseline mean error scores differed visually between groups (mean – SD degrees of error at pretest 2.0 – 0.6 and 1.5 – 0.5, respectively). Eils and Rosenbaum[33] concluded that position sense improved significantly after 6 weeks of sensorimotor training. Strikingly, the intervention group improved to the same level as the control group. It is possible that subjects in the control group performed best at pretest and no improvement was possible, whereas the intervention group left room for improvement after a poor performance at baseline. It is unknown whether this result is due to a training effect or a learning effect. From these studies it can be concluded that subjects’ improvements in angle reproduction after balance training is more likely attributable to a learning effect than to the effect of sensorimotor training. This conclusion also applies to subjects with a recent ankle sprain as well as subjects with ankle instability. ª 2009 Adis Data Information BV. All rights reserved.
2.2 Morphological Changes
Three studies have looked into the effect of sensorimotor training on strength.[27,35,36] Peak torques were measured by means of an isokinetic dynamometer. Kaminski et al.[36] evaluated eversion to inversion ratios (E/I ratios), which were calculated from average torque and peak torque measures. Given that the strength of the evertor muscles in individuals with chronic ankle instability had declined, a combination of sensorimotor training and strength exercises could be beneficial in preventing recurrent ankle sprains.[35,36] Holme et al.[27] chose another type of training with recently injured subjects. At 6 weeks and after 4 months, Holme et al.[27] found no significant differences in strength between training and control groups. There was a strength difference between the injured and the uninjured side 6 weeks after ankle injury. After 4 months, ankle strength was normalized. It is plausible that Sports Med 2009; 39 (7)
Sensorimotor Training for Ankle Injury Recovery
strength increase in the injured ankle was caused by natural recovery and was not evoked by sensorimotor training. Furthermore, pain in the ankle could be apparent in persons recovering from an ankle sprain, which could influence the maximal force deliverance during the tests. This could lead to a distorted impression of the amount of muscular force applied. Kaminski et al.[36] randomly assigned subjects to one of four groups: sensorimotor training, strength training, strength and sensorimotor training, and control. Baseline strength differences between groups were not described in detail. Results showed no significant increase in isokinetic E/I ratios and showed no significant differences between the four groups for E/I ratios at the end of the training protocol. Torque values in both directions were not presented, which makes it possible that strength increased proportionally in the eversion direction as well in the inversion direction. Kaminski et al.[36] stated that the method of testing was not specific enough to detect differences. The strength training protocol involved isotonic strength training exercises, but the tests were performed on an isokinetic dynamometer. Another theory is that the applied strength training was not vigorous enough and did not offer enough resistance to bring about changes in effect measures. Results of the study by Høiness et al.[35] revealed that in 19 subjects with chronic ankle instability the injured ankle was 11.5% weaker than the non-injured ankle. In contrast to Holme et al.,[27] pain did not hinder subjects’ performance. After sensorimotor training, subjects in the intervention group increased their peak eversion torque, whereas the control group showed no increase. At baseline testing no significant differences were found between the groups with respect to eversion torque. In contrast to Kaminski et al.[36] and Holme et al.,[27] Høiness et al.[35] found an increase in peak eversion torque. A possible shortcoming in the studies by Holme et al.[27] and Høiness et al.[35] is that average torque was not measured during subjects’ efforts. It is highly possible that not only the peak torques changed, but the average torque as well. ª 2009 Adis Data Information BV. All rights reserved.
599
The described results indicate that the effects of sensorimotor training and strength training on muscle torques are minimal, if present at all. The intensity of training needs to be high, either in frequency or in resistance of training, and the method of testing needs to be highly specific to the training method in order to demonstrate increases in strength. The question that rises is whether isokinetic measurement of muscle torques gives a realistic view of the strength that can be applied during sudden inversion, which probably does not elapse in a fully isokinetic way. 2.3 Combination of Neurophysiological and Morphological Changes
Eils and Rosenbaum[33] studied the effect of sensorimotor training on muscle reaction time after ankle inversion perturbations. The interpretation of study results is limited, because a thorough analysis of baseline muscle reaction time differences between groups was not carried out. Muscle reaction times for peroneus longus and peroneus brevis were significantly prolonged after 6 weeks in both groups. As a result of this prolonged reaction, the activity of the peroneus muscles will synchronize with the activity pattern of the tibialis anterior muscle. Through this co-contraction, a higher joint stiffness could be established, which may have led to an increased ability to resist movements. Prolongation of reaction time is counterintuitive to an increased ability to resist movements. However, it is possible that other strategies for an optimized reaction were utilized. Differences between pre- and posttests for the Integrated Electromyography were not found,[33] which might imply that no increase in muscular strength was present directly after sudden inversion. The question is raised whether the minimal reaction time changes in the study by Eils and Rosenbaum[33] make any difference during the reflex reaction of muscles after sudden inversion. From previous research[58,59] it can be concluded that a sudden inversion lasts 40–45 msec. Reaction times measured in the study by Eils and Rosenbaum[33] are too long to sort effect on the Sports Med 2009; 39 (7)
Hupperets et al.
600
degree of inversion, even after the training period (reaction times of 60–70 msec). 2.4 Functional Changes
The majority of studies[27,32-35,37-39,44,51] focused on the effects of sensorimotor training on static postural sway, whereas one study[50] measured dynamic postural stability. Postural sway was determined as a measure of balance on a force platform[27,32,33,37-39,44,50,51] or a balance system[34,35] in a one-legged stance. Two studies investigated the effect of a sensorimotor training programme on postural sway in subjects with an ankle sprain in the previous year.[27,39] Holme et al.[27] found no significant differences in postural sway between the two groups after 6 weeks and 4 months. After 6 weeks there were no differences between affected and non-affected ankles in both groups, whereas after 4 months a difference was found. This would suggest that balance improvements were due to natural recovery and not due to sensorimotor training. Although a difference in the number of positive anterior drawer signs was reported, it is unlikely that this affected results. Verhagen et al.[39] randomly assigned 22 subjects to either an intervention or a control group. An additional eight subjects were participants in an organized volleyball competition and were assigned to a convenience volleyball group. All subjects in the intervention group and the volleyball group received a 5.5-week sensorimotor training programme, whereas subjects in the control group received no training. At baseline no significant postural sway differences were found between all groups. Postural sway improved equally in all three groups after 5.5 weeks. This improvement was thought to be due to a learning effect and not to sensorimotor training. Other studies reported on the effect of sensorimotor training on postural sway in subjects with ankle instability.[32-35,37,38,44,50,51] Bernier and Perrin[32] focused primarily on sensorimotor training, where others offered a combination of both sensorimotor training and strength exercises.[33-35,37,38,44,50,51] Bernier and Perrin[32] tested subjects on a moving platform as well as on ª 2009 Adis Data Information BV. All rights reserved.
a static one, both with eyes open and with eyes closed. Results indicated an improvement in proprioception, since the training group had a significant postural sway and balance improvement in the stable platform condition (eyes closed) and the moving platform condition (eyes open). The condition in which subjects’ postural sway and balance was assessed on a stable platform with eyes open showed a trend in improvement of the experimental group. Improvements in the eyes closed condition might be attributable to an enhanced proprioception, since subjects cannot rely on their visual system in that situation. However, a learning effect of repeated testing is more likely, because after 6 weeks the control group showed reduced postural sway in both the eyes open and the eyes closed conditions compared with baseline. Powers et al.[38] used the same strength and training protocol as Kaminski et al.,[36] showing no differences in static balance in all groups after the training period. The fact that only one trial was used to measure balance at pretest and at post-test makes it difficult to interpret results. Although this choice was made to rule out a learning or practice effect and the effect of fatigue,[38] it is recommended to use multiple trials because of susceptibility to faulty measurements. Whereas a study duration of 6 weeks is considered to be a gold standard among various studies,[32-35,37,38,50,51] Hess and colleagues[34] prescribed an exercise programme of only 4 weeks. They found no significant differences on sway in any direction between groups after 4 weeks. This could be because the duration of the training programme was too short to bring about functional changes. Another reason is that with the chosen type of training (emphasis on agility) the dynamic balance was improved, which may not show in a static balance test. The fact that differences were found in subject characteristics at baseline does not alter the conclusion drawn from the study. Hale et al.[44] randomly assigned 29 subjects with chronic ankle instability to either an intervention or a control group. Another 19 subjects without chronic instability were assigned to a healthy group, but are not of interest in this review. Sports Med 2009; 39 (7)
Sensorimotor Training for Ankle Injury Recovery
Subjects in the intervention group received a 4-week sensorimotor training programme, whereas subjects in the control group received no training. At baseline and at 4 weeks’ follow-up, no significant postural sway differences were found between groups. Possible reasons for not finding an improved postural control were: (i) discarding and repeating trials at baseline; and (ii) the intensity and focus of the training programme.[44] Hale et al.[44] also measured functional improvements through the Star Excursion Balance Test (SEBT) and self-reported function through the Foot and Ankle Disability Index (FADI and FADI-Sport). At baseline there were no differences in SEBT and FADI between intervention and control groups. The mean reach distance of the SEBT intervention group improved significantly after sensorimotor training compared with controls. This change was apparent in the posteromedial, posterolateral and lateral directions, as well as in the mean of all eight reach directions. The intervention group showed more improvement in FADI scores at follow-up compared with the control group. As discussed by Hale et al.,[44] this improvement could possibly be attributed to a placebo effect, since the control group did not participate in any rehabilitation programme. In the study by Høiness et al.,[35] both groups improved in single-legged stance after training. Compared with baseline testing, the control group showed more improvement than the training group did after 6 weeks. However, there was already a difference in postural sway between groups at baseline. A missing definition for history of injury hampers comparability with other relevant studies, but has no effect on the conclusion that the sensorimotor training was not effective. It is more likely that the improved single-legged stance was due to a learning effect. The majority of studies[32-34,37,38,50] lack a clear analysis of postural sway differences between training and control groups at baseline testing. Baseline outcome measures data were presented as means and standard deviations only. This makes it difficult to assess whether statistical analyses had been undertaken and to interpret the impact of possible significant differences between groups at the end of a training programme. ª 2009 Adis Data Information BV. All rights reserved.
601
Eils and Rosenbaum[33] showed an improvement for all postural sway parameters in both the intervention group and the control group after 6 weeks. Furthermore, the intervention group improved on static balance in the mediolateral direction,[33] while the control group showed no improvement. However, the control group showed an improvement in static balance in the anteroposterior direction. A nonsignificant reduction of sway in that direction was found in the training group. Eils and Rosenbaum[33] attributed these results to a short-term adaptation by a learning process in the control group. A long-term adaptation might apply for the intervention group as a result of the exercise programme. This conclusion needs to be put in the right perspective, since it is partly based on statistically nonsignificant findings. Kidgell et al.[37] found an effect of a balance training programme on postural stability. After 6 weeks the training groups (dura disc and minitrampoline) showed a significant improvement in postural sway compared with the control group. However, these results need to be looked at with caution because of the small observed power (0.233). This small power was due to the small sample size that was chosen (n = 20).[37] The mentioned difference in body height between groups might have influenced results, but this is not probable. Ross and Guskiewicz[50] studied the effects of a 6-week sensorimotor training programme with and without stochastic resonance (SR) stimulation on dynamic postural stability in subjects with functional ankle instability by measuring timeto-stabilization (TTS). Thirty subjects were assigned to a conventional training group (CCT), SR stimulation training group (SCT), or a control group. The training programme led to an improved anterior/posterior and medial/lateral TTS compared with baseline for both training groups. However, compared with the control group, this improvement was not significant. It is striking that pretest scores of anterior/posterior TTS differed visually between CCT and control (pretest mean TTS – SD: 2.2 – 0.8 and 1.6 – 0.4, respectively). Ross et al.[51] used the same training protocol as Ross and Guskiewicz.[50] Thirty subjects with Sports Med 2009; 39 (7)
Hupperets et al.
602
functional ankle instability were assigned to the same three groups: CCT, SCT or control group. Whereas Ross and Guskiewicz[50] focused on dynamic postural stability, Ross et al.[51] measured static postural stability. Group differences were not present at baseline. After 6 weeks no difference in postural sway was found between CCT and control, but the SCT group had reduced post-test means compared with pooled post-test means of CCT and control. From these findings it can be concluded that sensorimotor training alone did not result in significantly better postural stability than in subjects who did not participate in sensorimotor training. SR stimulation might be used as an alternative therapy to improve postural stability deficits associated with functional ankle instability. From these studies the conclusion can be drawn that an effect of sensorimotor training on postural sway was masked by a learning effect of repeated measures. Whether sensorimotor training led to a short-term improvement in postural sway is not shown in these studies. A follow-up period longer than 6 weeks seems necessary to address long-term effects of sensorimotor training on postural regulation. 3. Training versus Learning Effect The goal of this critical review was to determine the effects of sensorimotor training on morphological and neurophysiological properties of the previously sprained ankle on the one hand and functional characteristics on the other hand. Based on the discussed studies, no effect of sensorimotor training on ankle characteristics was found. This applies to morphological and neurophysiological ankle properties, as well as to functional characteristics. In our opinion, changes in morphology and neurophysiology (e.g. improved joint position sense, decreased muscle reaction times and increased strength) induced by sensorimotor training lead to functional improvements of the ankle, which subsequently lead to reduced re-injury risk (see figure 1). Possible neurophysiological adaptations are changes in physiological processes such as the sensory threshold of specific peripheral mechanoª 2009 Adis Data Information BV. All rights reserved.
receptors, nerve conduction velocity, sensorimotor integration at the spinal and/or supraspinal level, alpha motoneuron pool excitability and/or gamma-motor neuron/muscle spindle function. Potential changes in morphology are changes in muscle cross-sectional area, myofibril size and/or ligament structure. Hertel[31] validates our conceptual model in stating that postural control deficits are likely due to a combination of impaired proprioception and neuromuscular control. Hertel’s use of the term ‘functional insufficiencies’ was developed to show that individual authors were using several different sensorimotor constructs (proprioception, postural control, strength and neuromuscular responses to inversion perturbation) to indentify the condition of functional ankle instability. Although it was not meant to state that proprioceptive measures were truly functional, multiple studies on this topic have blindly copied Hertel’s view in determining effects of sensorimotor training, without making a distinction between properties, which in our opinion is wrong. The first aim in defining the underlying mechanisms responsible for the preventive effect of sensorimotor training in subjects with previous ankle sprains is establishing the effects on morphological and neurophysiological characteristics of the ankle (e.g. joint position sense, muscle reaction times and strength). From the literature reviewed, it can be concluded that improvements in position sense and reaction times are merely caused by a learning process of repeated testing and are not a training effect. This might imply a Sensorimotor training
Morphological changes
Neurophysiological changes
Mechanical changes
Functional changes
Re-injury risk Fig. 1. Conceptual model of the pathway of proprioceptive training leading to a reduced re-injury risk of the ankle.
Sports Med 2009; 39 (7)
Sensorimotor Training for Ankle Injury Recovery
validity problem, while no change was to be expected in controls. As mentioned by AshtonMiller et al.,[14] postural sway measurements are skills in motor tasks and are no true measures of proprioception. As direct measurement of proprioception is impossible, these skills in motor tasks are the closest means to measuring proprioception. The future focus needs to be on establishing methods more closely linked to proprioception. A new potential workable method is TTS, which is a functional measure of stability. This measure of dynamic stability forces subjects to maintain balance through a transition from a dynamic to a static state.[60] Effects on strength do not seem to be apparent. An important observation by Holme et al.[27] was that at 12-month follow-up there was a significant group difference in re-injury rate. Despite finding no significant improvements in postural sway, joint position sense and strength after training, a secondary preventive effect was found. In general, a relatively low number of on-topic and relevant studies were found. In addition, a differentiation had to be made between studies examining interventions in patients after acute ankle sprains compared with patients with ankle instability. Furthermore, an important factor is the details of the rehabilitation protocols. An adequate training stimulus was not always provided to subjects, which could have influenced results. The majority of studies scored negative on homogeneity of the intervention and control group with regard to subject characteristics. This is likely due to the low number of subjects, which is typical for these types of laboratory-based studies. Study populations of the studies discussed here ranged from 19 to 92 subjects. Studies must rely on larger sample sizes for higher statistical power to be able to detect smaller differences in outcome measures. A multidisciplinary approach, combining biomechanics and epidemiology, preferably using a randomized controlled trial design, is recommended to obviate this issue of small sample size. To date no study has been conducted on the effect of sensorimotor training on the mechanical aspect of recurrent ankle sprains. Mechanical instability of the ankle complex occurs as a result ª 2009 Adis Data Information BV. All rights reserved.
603
of morphological changes after an initial ankle sprain. These changes may include pathological laxity, impaired arthrokinematics, synovial changes and the development of degenerative joint disease, which may all occur in combination or in isolation.[31] Chronic ankle instability, increasing re-injury risk, in subjects may not only be caused by mechanical instability or functional instability alone, but by a combination of these two entities.[61,62] It could be that sensorimotor training affects those morphological factors closely linked to mechanical stability. However, this has never been studied, while functional changes induced by sensorimotor training seem more plausible. 4. Conclusion The pathway through which sensorimotor training reduces re-injury risk remains unclear. The ‘black box’ on the ‘true’ effect of sensorimotor training remains unopened. Using enhanced measurement techniques, which equate to specific physiological processes that are inside the ‘black box’, would be beneficial for future research. To create more insight in the pathway reducing re-injury risk, studies should: (i) differentiate between morphological, physiological and functional changes; (ii) use larger sample sizes with a priori sample size calculations; (iii) ensure correspondence between training and test method; (iv) use other measures than postural sway more closely linked to functional stability; and (v) use a longer follow-up period than 6 weeks. Acknowledgements No sources of funding were used to assist in the preparation of this review. The authors have no conflicts of interest that are directly relevant to the content of this review.
References 1. Garrick JG, Requa RK. Role of external support in the prevention of ankle sprains. Med Sci Sports 1973; 5: 200-3
Sports Med 2009; 39 (7)
604
2. Rovere GD, Clarke TJ, Yates CS, et al. Retrospective comparison of taping and ankle stabilizers in preventing ankle injuries. Am J Sports Med 1998; 16: 228-33 3. Sitler M, Ryan J, Wheeler B, et al. The efficacy of a semi rigid ankle stabilizer to reduce acute ankle injuries in basketball: a randomized clinical study at West Point. Am J Sports Med 1994; 22: 454-61 4. Surve I, Schwellnuss MP, Noakes T, et al. A fivefold reduction in the incidence of recurrent ankle sprains in soccer players using the sport-stirrup orthosis. Am J Sports Med 1994; 22: 601-6 5. Wexler FK. The injured ankle. Am Fam Phys 1998; 57: 474-80 6. Tropp H, Ekstrand J, Gilquist J. Factors affecting stabilometry recordings of single limb stance. Am J Sports Med 1984; 12 (3): 185-8 7. Bahr R, Bahr IA. Incidence of acute volleyball injuries: a prospective cohort study of injury mechanisms and risk factors. Scand J Med Sci Sports 1997; 7: 166-71 8. Ekstrand J, Tropp H. The incidence of ankle sprains in soccer. Foot Ankle 1990; 11: 41-4 9. Milgrom C, Shlamkovitch N, Finestone A, et al. Risk factors for lateral ankle sprains: a prospective study among military recruits. Foot Ankle 1991; 12: 26-30 10. Verhagen EALM, van der Beek AJ, Bouter LM, et al. A one-season prospective cohort study of volleyball injuries. Br J Sports Med 2004; 38 (4): 477-81 11. Freeman MAR. Instability to the foot after injuries to the lateral ligament of the ankle. J Bone Joint Surg Br 1965; 47: 669-77 12. Karlsson J, Lansinger O. Chronic lateral instability of the ankle in athletes. Sports Med 1993; 16 (5): 355-65 13. Lephart SM, Conley K. The role of proprioception in chronic ankle instability. In: Schmidt R, Benesch S, Lipke K, editors. Chronic ankle instability: manuscripts of the International Ankle Symposium. Ulm: Libri Verlag, 2000: 254-67 14. Ashton-Miller JA, Wojtys EM, Huston LJ, et al. Can proprioception really be improved by exercises? Knee Surg Sports Traumatol Arthrosc 2001; 9: 128-36 15. Taube W, Gruber M, Gollhofer A. Spinal and supraspinal adaptations associated with balance training and their functional relevance. Acta Physiol 2008; 193 (2): 101-16 16. Paterno MV, Myer GD, Ford KR, et al. Neuromuscular training improves single-limb stability in young female athletes. J Orthop Sports Phys Ther 2004; 34: 305-16 17. Banaschewski T, Besmens F, Zieger H, et al. Evaluation of sensorimotor training in children with ADHD. Percept Mot Skills 2001; 92: 137-49 18. Gruber M, Gollhofer A. Impact of sensorimotor training on the rate of force development and neural activation. Eur J Appl Physiol 2004; 92: 98-105 19. Tropp H, Askling C, Gillquist J. Prevention of ankle sprains. Am J Sports Med 1985; 13 (4): 259-62 20. Olsen OE, Myklebust G, Engebretsen L, et al. Exercises to prevent lower limb injuries in youth sports: cluster randomised controlled trial. Br Med J 2005; 330: 449-52 21. Petersen W, Hohmann G, Pufe T, et al. A controlled prospective case control study of a prevention training program in female team handball players: the German experience. Arch Orthop Trauma Surg 2005; 125 (9): 614-21
ª 2009 Adis Data Information BV. All rights reserved.
Hupperets et al.
22. Wedderkopp N, Kaltoft M, Lundgaard B, et al. Prevention of injuries in young female players in European team handball: a prospective intervention study. Scand J Med Sci Sports 1999; 9 (1): 41-7 23. Wedderkopp N, Kaltoft M, Holm R, et al. Comparison of two intervention programmes in young female players in European team handball – with and without ankle disc. Scand J Med Sci Sports 2003; 13 (6): 371-5 24. Bahr R, Lian O, Bahr IA. A twofold reduction in the incidence of acute ankle sprains in volleyball after the introduction of an injury prevention program: a prospective cohort study. Scand J Med Sci Sports 1997; 7 (3): 172-7 25. Stasinopoulos D. Comparison of three preventive methods in order to reduce the incidence of ankle inversion sprains among female volleyball players. Br J Sports Med 2004; 38 (2): 182-5 26. Verhagen E, van der Beek AJ, Bouter LM, et al. The effect of a proprioceptive balance board training program for the prevention of ankle sprains: a prospective controlled trial. Am J Sports Med 2004; 32 (6): 1385-93 27. Holme E, Magnusson SP, Becher K, et al. The effect of supervised rehabilitation on strength, postural sway, position sense and re-injury risk after acute ankle ligament sprain. Scand J Med Sci Sports 1999; 9 (2): 104-9 28. McGuine TA, Keene JS. The effect of a balance training program on the risk of ankle sprains in high school athletes. Am J Sports Med 2006; 34 (7): 1103-11 29. Mohammadi F. Comparison of 3 preventive methods to reduce the recurrence of ankle inversion sprains in male soccer players. Am J Sports Med 2007; 35 (6): 922-6 30. Eils E. The role of proprioception in the primary prevention of ankle sprains in athletes: a review. Int Sports Med J 2003; 4: 1-9 31. Hertel J. Functional anatomy, pathomechanics, and pathophysiology of lateral ankle instability. J Athlet Train 2002; 37 (4): 364-75 32. Bernier JN, Perrin DH. Effect of coordination training on proprioception of the functionally unstable ankle. J Orthop Sports Phys Ther 1998; 27 (4): 264-75 33. Eils E, Rosenbaum D. A multi-station proprioceptive exercise program in patients with ankle instability. Med Sci Sports Exerc 2001; 33 (12): 1991-8 34. Hess DM, Joyce CJ, Arnold BL, et al. Effect of a 4-week agility training program on postural sway in the functionally unstable ankle. J Sports Rehab 2001; 10: 24-35 35. Høiness P, Glott T, Ingjer F. High-intensity training with a bi-directional bicycle pedal improves performance in mechanically unstable ankles: a prospective randomized study of 19 subjects. Scand J Med Sci Sports 2003; 13 (4): 266-71 36. Kaminski TW, Buckley BD, Powers ME, et al. Effect of strength and proprioceptive training on eversion and inversion strength ratios in subjects with unilateral functional ankle instability. Br J Sports Med 2003; 37 (5): 410-5 37. Kidgell DJ, Horvath DM, Jackson BM, et al. Effect of six weeks of dura disc and mini-trampoline balance training on postural sway in athletes with functional ankle instability. J Strength Cond Res 2007; 21 (2): 466-9
Sports Med 2009; 39 (7)
Sensorimotor Training for Ankle Injury Recovery
38. Powers ME, Buckley BD, Kaminski TW, et al. Six weeks of strength and proprioception training does not affect muscle fatigue and static balance in functional ankle instability. J Sport Rehab 2004; 13: 201-27 39. Verhagen E, Bobbert M, Inklaar M, et al. The effect of a balance training programme on centre of pressure excursion in one-leg stance. Clin Biomech 2005; 20 (10): 1094-100 40. Riemann BL. Is there a link between chronic ankle instability and postural instability? J Athl Train 2002; 37 (4): 386-93 41. Akhbari B, Ebrahimi Takamjani I, Salavati M, et al. Ankle musculature latency measurement to varying angles of sudden external oblique perturbation in normal functionally unstable ankles. Med J Islamic Rep Iran 2007; 20: 166-74 42. Docherty CL, Valovich McLeod TC, Shultz SJ. Postural control deficits in participants with functional ankle instability as measured by the balance error scoring system. Clin J Sport Med 2006; 16: 203-8 43. Gauffin H, Tropp H, Odenrick P. Effect of ankle disk training on postural control in patients with functional instability of the ankle joint. Int J Sport Med 1988; 9: 141-4 44. Hale SA, Hertel J, Olmsted-Kramer LC. The effect of a 4-week comprehensive rehabilitation program on postural control and lower extremity function in individuals with chronic ankle instability. J Orthop Sports Phys Ther 2007; 37 (6): 303-11 45. Kaminski TW, Perrin DH, Gansneder BM. Eversion strength analysis of uninjured and functionally unstable ankles. J Athl Train 1999; 34: 239-45 46. Matsusaka N, Yokoyama S, Tsurusaki T, et al. Effect of ankle disk training combined with tactile stimulation to the leg and foot on functional instability of the ankle. Am J Sports Med 2001; 29: 25-30 47. Michell TB, Ross SE, Blackburn JT, et al. Functional balance training, with or without exercise sandals, for subjects with stable or unstable ankles. J Athl Train 2006; 41: 393-8 48. Osborne MD, Chou LS, Laskowski ER, et al. The effect of ankle disk training on muscle reaction time in subjects with a history of ankle sprain. Am J Sports Med 2001; 29: 627-32 49. Pintsaar A, Brynhildsen J, Tropp H. Postural corrections after standardised perturbations of single limb stance: effect of training and orthotic devices in patients with ankle instability. Br J Sports Med 1996; 30: 151-5 50. Ross SE, Guskiewicz KM. Effect of coordination training with and without stochastic resonance stimulation on dynamic postural stability of subjects with functional ankle instability and subjects with stable ankles. Clin J Sport Med 2006; 16 (4): 323-8
ª 2009 Adis Data Information BV. All rights reserved.
605
51. Ross SE, Arnold BL, Blackburn JT, et al. Enhanced balance associated with coordination training with stochastic resonance stimulation in subjects with functional ankle instability: an experimental trial. J Neuroeng Rehabil 2007 Dec 17; 4: 47 52. Rozzi SL, Lephart SM, Sterner R, et al. Balance training for persons with functionally unstable ankles. J Orthop Sports Phys Ther 1999; 29: 478-86 53. Sekir U, Yildiz Y, Hazneci B, et al. Effect of isokinetic training on strength, functionality and proprioception in athletes with functional ankle instability. Knee Surg Sports Traumatol Athrosc 2007; 15: 654-64 54. Sheth P, Yu B, Laskowski ER, et al. Ankle disk training influences reaction times of selected muscles in a simulated ankle sprain. Am J Sports Med 1997; 25: 538-43 55. Willems T, Witvrouw E, Verstuyft J, et al. Proprioception and muscle strength in subjects with a history of ankle sprains and chronic instability. J Athl Train 2002; 37: 487-93 56. Verhagen EA, van Mechelen W, de Vente W. The effect of preventive measures on the incidence of ankle sprains. Clin J Sports Med 2000; 10 (4); 291-6 57. Borghouts JAJ, Koes BW, Bouter LM. The clinical factors of non-specific neck-pain: a systematic review. Pain 1998; 77: 1-13 58. Ashton-Miller JA, Ottaviani RA, Hutchinson C, et al. What best protects the inverted weightbearing ankle against further inversions? Am J Sports Med 1996; 24: 800-9 59. Gruneberg C, Nieuwenhuijzen PHJA, Duysens J. Reflex responses in the lower leg following landing impact on an inverting and non-inverting platform. J Physiol 2003; 550: 985-93 60. Goldie P, Bach T, Evans O. Force platform measures for evaluating postural control: reliability and validity. Arch Phys Med Rehab 1989; 70: 510-7 61. Tropp H, Odenrick P, Gillquist J. Stabilometry recordings in functional and mechanical instability of the ankle joint. Int J Sports Med 1985; 6: 180-2 62. Wilkerson GB, Nitz AJ. Dynamic ankle stability: mechanical and neuromuscular interrelationships. J Sport Rehabil 1994; 3: 43-57
Correspondence: Drs Maarten D.W. Hupperets, Department of Public and Occupational Health, Institute for Research in Extramural Medicine, VU University Medical Centre, Van der Boechorststraat 7, 1081 BT, Amsterdam, the Netherlands. E-mail:
[email protected]
Sports Med 2009; 39 (7)
