Journal of Sports Sciences, 2004, 22, 629–636
Groin injuries in professional rugby league players: a prospective study DONNA M. O’CONNOR* Human Movement and Health Education, Faculty of Education, University of Sydney, Sydney, NSW 2006, Australia
Accepted 3 December 2003
Altogether, 100 uninjured professional rugby league players were evaluated over a 2-year period. Their height, body mass, sum of skinfolds, girths and bone diameters were recorded. A Cybex 340 isokinetic dynamometer was used to determine peak torque, work, power, endurance ratios and peak torque ratios of the hip abductors and adductors (5 repetitions at 0.52 and 2.08 rad × s7 1; 20 repetitions at 3.66 rad × s7 1) and knee flexors and extensors (4 repetitions at 1.04 and 3.14 rad × s7 1; 30 repetitions at 5.22 rad × s7 1). Hip abduction and adduction were also assessed with the hip in external rotation. Discriminant function analysis was conducted on all predictor variables to develop a multivariate predictive model capable of classifying players with a high degree of accuracy into groups with and without a groin injury. The model consisted of eight variables and correctly classified 91.7% of the non-injured players and 90.5% of the injured players. The correct classification for the model as a whole was 91.4%. The aetiological factors identified as being related to injury of the groin musculotendinous unit included abduction and adduction-with-rotation peak torque, angle of adduction and abduction-with-rotation peak torque, strength ratio of hip muscle groups, bilateral difference in extension peak torque, femur diameter and body mass. Keywords: discriminant analysis, groin injuries, isokinetic, rugby league.
Introduction The prevalence of groin injuries in athletes is recognized as one of the most difficult problems in sport (Renstrom, 1992). Groin pain is a common and debilitating injury in all football codes. According to Estwanik et al. (1990) and Morelli and Smith (2001), groin injuries account for 2–5% of all sports injuries. Ekstrand (1994), however, stated that an injury rate of up to 28% had been reported for footballers with groin pain. The increased incidence of groin injuries among rugby league players is of concern to coaches. Gibbs (1993) reported that medial collateral knee ligament sprains and groin musculotendinous strains were the most common specific injuries sustained by rugby league players (each 10.6%). Seven of the 15 players with groin strains eventually required surgery (conjoint tendon repair or adductor tenotomy) with a complete recovery within 8–12 weeks of surgery. Estwanik et al. (1990) maintained that groin strains frequently occur in sports that involve quick accelera* Address all correspondence to e-mail: d.o’
[email protected]. edu.au
tion and sudden changes in direction due to the abnormal lengthening of the muscle components surrounding the hip. Further analyses of groin injuries (Smodlaka, 1980; Kulund, 1982) suggest that sudden, powerful over-stretching of the leg and thigh in abduction and external rotation is the cause of such injuries. This view supports that of Merrifield and Cowan (1973), who reported that it is the strain of the adductor muscles caused by a sudden violent external rotation of the thigh while the leg is widely abducted and the foot firmly planted (e.g. during a side-step manoeuvre) that leads to groin pain. It has been suggested that many factors, rather than a single variable, may play a role in the development of injury (Lysens et al., 1991; Meeuwisse, 1994a,b; Schootman et al., 1994). Several aetiological factors have been proposed as being related to injury of the groin musculotendinous unit, including muscle tightness, strength or flexibility asymmetry; insufficient strength of prime muscle groups; fatigue; incorrect technique; playing conditions; training errors; inadequate warm-up; and too early a return to competition following an injury (Merrifield and Cowan, 1973; Poulmedis, 1985; Fried and Lloyd, 1992; Lynch and Renstrom, 1999; Emery and Meeuwisse, 2001; Tyler
Journal of Sports Sciences ISSN 0264-0414 print/ISSN 1466-447X online # 2004 Taylor & Francis Ltd DOI: 10.1080/02640410310001655804
630 et al., 2001; Nicholas and Tyler, 2002). Several researchers have attempted to determine risk factors associated with groin injuries in ice hockey players. A strength and power imbalance of greater than 25% and a hip abduction range of motion deficit of 48 were identified in players who were prone to adductor strains (Merrifield and Cowan, 1973). More recently, Tyler et al. (2001) reported that players with weaker adductor muscles had a predisposition to an adductor strain. Furthermore, a player with an adductor-toabductor strength ratio of less than 80% was 17 times more likely to sustain an adductor injury. This contradicts the findings of Emery and Meeuwisse (2001), who indicated that previous injury and insufficient sport-specific training during the offseason rather than peak isometric adductor strength and abduction flexibility were predictors of groin injury. There is a need to identify and quantify the causes of groin injuries in rugby league to help determine who is at risk of sustaining such an injury, and then to employ methods and techniques to minimize the number of potential victims. By studying the occurrence of groin injuries and establishing the aetiological factors and mechanisms, trends can be identified that may help in developing strategies to reduce the number or severity of future groin injuries. There has been much conjecture about mechanisms associated with musculotendinous groin injuries. The aetiology and distribution of groin injuries must be ascertained before any prophylactic recommendations can be proposed. The aim of this study was to develop a predictive model that could accurately classify players that sustained a groin injury.
Methods Participants In total, 100 professional rugby league players (58 backs, 42 forwards) from 13 clubs involved in the Australian Rugby League first or reserve grade competition volunteered to participate in the study after a visit to a club training session by the researcher. The players had never sustained a groin injury prior to the study. Written informed consent was received from all players and ethical approval was granted by the University of Sydney Human Ethics Research Committee. Descriptive characteristics of the participants are outlined in Table 1. Procedures All assessments were conducted by the same researcher, although three different facilities were used to accommodate teams from different cities. A questionnaire was
O’Connor Table 1. Descriptive characteristics of the players (means + s)
Age (years) Height (m) Body mass (kg) Skinfolds (mm) Playing experience (years)
Injured
Non-injured
22.2+2.9 1.80+0.13 90.5+9.5 75.2+20.7 3.7+2.5
20.2+4.5 * 1.78+0.06 84.7+10.2 * 70.6+19.5 2.9+1.9
* P 5 0.05.
distributed to players to ascertain age, playing position, experience in grade football, level of ability, weight training experience and emphasis placed on leg and hip muscles in their weights programme. Kinanthropometric measures included height, body mass, sum of eight skinfolds (tricep, subscapular, biceps, iliac crest, supraspinale, abdominal, thigh, calf), girths (thigh, hip) and bone diameters (femur, bi-iliac) (for protocols, see Norton and Olds, 1996). Hip abduction, adduction, internal and external hip rotation range of motion and hamstring and groin flexibility (Faber test and splits) were assessed (for protocols, see Norkin and White, 1985). The order of concentric/concentric isokinetic testing was knee flexion–extension of each limb, followed by abduction–adduction, and then abduction–adduction with external rotation using a Cybex 340 dynamometer (Lumex Inc., USA). Angular velocities were chosen from those adopted in previous literature (Poulmedis, 1985; Cahalan et al., 1989; Donatelli et al., 1991; Kushner et al., 1992; Reilly and Florida-James, 1995) and supported by recommendations from the isokinetic manual. Test–retest correlation coefficients of 0.75– 0.83 for hip abduction/adduction were recorded in a pilot study. Calibration of the dynamometer was automatically administered at the beginning of each test session. The leg to be tested first was determined randomly. Throughout all testing with the isokinetic dynamometer, the participants were verbally encouraged to perform maximal contractions through their full range of motion. The participants were also instructed to stop the test if they felt any unusual pain or discomfort during the procedure. Knee flexion–extension The participants were seated with the axis of the dynamometer aligned with the lateral condyle of the femur. The tibial pad was placed at the distal third of the tibia and attached by a velcro strap. The thigh, torso and non-involved leg were all stabilized using velcro straps. The angle between the back support and seat of the Cybex chair was 100o. Each participant was
631
A predictive model of groin injuries instructed to hold onto the handgrips. Gravity correction was performed with the knee in full extension. A four-repetition submaximal warm-up for familiarization was performed at 1.04, 3.14 and 5.22 rad × s7 1 immediately before each test set. The test protocol consisted of four maximal knee flexion–extension contractions at 1.04 rad × s7 1 over 15 s. After a 60-s rest interval, four repetitions were conducted over 6 s with the dynamometer set to 3.14 rad × s7 1. Finally, at 5.22 rad × s7 1 a set of 30 maximal repetitions was performed over 30 s. The full procedure was then administered to the opposite leg after 3 min of recovery. Hip abduction–adduction The player was positioned on his side (facing away from the Cybex machine) on the upper body extension table with the axis of the dynamometer aligned orthogonally with the greater trochanter. The thigh pad was attached proximally to the knee. The torso, pelvis and contralateral thigh were stabilized with velcro straps with the iliac crest aligned in the coronal plane to eliminate any rotation. The player’s head rested on the table extension pad. The top hand crossed the body to grasp the long stabilization handle. After gravity correction, each player was given four warm-up repetitions at each angular velocity (0.52, 2.08 and 3.66 rad × s7 1) to familiarize himself with the movement speed. Five repetitions were completed at angular velocities of 0.52 and 2.08 rad × s7 1. The Cybex was then set at 3.66 rad × s7 1 with 20 maximal repetitions being performed. The players were allowed 60 s recovery between each test velocity. The hips were kept in neutral so the movement was in the coronal plane. The knee under study was extended throughout testing. The participants were then given 3 min rest, during which time they were repositioned for testing of the contralateral limb.
was recorded if there was: (a) pain and tenderness in the adductors or at the adductor bone–tendon junction; (b) pain and weakness on resisted adduction; and (c) the player was unable to complete training or a game and missed the following training session. Statistical analysis Discriminant function analysis was conducted on all predictor variables to develop a prospective multivariate predictive model for the groin-injured and non-injured groups. The block method rather than the stepwise method was selected, as it was argued that the block method would lead to an increasingly replicable classification of players. The analyses were performed using the SPSS for Windows software.
Results Over the 2-year period, the risk of groin injury in rugby league players was high (23%) and similar to the rate reported by Gibbs (1993). In the present study, the injuries documented were moderate (5–16 weeks duration) to severe (requiring surgery) in nature. Seventy percent of injured players experienced pain for at least 7 weeks, with three players undergoing adductor tenotomy surgery. Exposure to injury was estimated using training and game time and was calculated to be 2.4 per 1000 hours. The means and standard deviations for hip abduction and adduction at the three angular velocities are presented in Table 2. The hip adductors always produced a greater torque than the hip abductors at each of the three test speeds for both the injured and noninjured groups. The non-injured players generated a significantly higher peak torque in adduction with the non-dominant limb at 0.52 and 2.08 rad × s7 1 (P 50.05). There was no significant difference between groups for knee flexion–extension.
Hip abduction–adduction with external rotation The hip abduction–adduction protocol as outlined above was repeated, but this time the hip was stabilized in neutral with external rotation. The player was encouraged to keep the knee extended but point the toes to the ceiling. The iliac crest was still in the coronal plane. Maximal repetitions were conducted at 0.52, 2.08 and 3.66 rad × s7 1 for each limb. Diagnosis of groin injury Over a 2-year period, the players were monitored closely for diagnosis of a groin injury by the club physician or physiotherapist. When an injury occurred, mechanism, site and severity were recorded. An injury
Discriminant function analysis Discriminant function analysis identified eight predictor variables that resulted in the best categorization of the injured and non-injured. The abbreviations used to denote each variable are presented in Table 3. The variables angle of peak torque during abduction with external rotation using the dominant limb at 2.08 rad × s7 1, peak torque during abduction at 0.52 rad × s7 1 with the non-dominant limb and dominant femur diameter recorded the lowest lambda values, indicating the means of these variables differed the most between non-injured and injured footballers (Table 4). Intercorrelations among predictor variables are summarized in Table 5. The intercorrelations for these
632
O’Connor
Table 2. Means and standard deviations for isokinetic data (N × m unless otherwise stated)
Femur diameter, dominant (cm) Abduction PT, dominant Abduction PT, non-dominant Adduction PT, dominant Adduction PT, non-dominant Abduction PT, dominant Abduction PT, non-dominant Adduction PT, dominant Adduction PT, non-dominant Abduction PT, dominant Abduction PT, non-dominant Adduction PT, dominant Adduction PT, non-dominant Angle PT add , dominant (8) Angle PT add, non-dominant (8) Add/rot PT, dominant Add/rot PT, non-dominant Knee extension PT, dominant Knee extension PT, non-dominant Knee flexion PT, dominant Knee flexion PT, non-dominant PT bilateral deficit knee extension (%)
0.52 0.52 0.52 0.52 2.08 2.08 2.08 2.08 3.66 3.66 3.66 3.66 3.66 3.66 0.52 0.52 1.04 1.04 1.04 1.04 3.14
rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71 rad × s71
Non-injured
Injured
9.63 + 0.4 152.7 + 36.6 153.3+36.5 168.1+72.2 163.7+61.5 114.3+34.7 119.5+30.4 150.8+67.8 146.3+65.7 65.0+33.4 64.9+37.2 90.8+57.3 88.2+54.2 23.2+13.4 24.6+11.8 180.1+74.8 184.1+58.5 262.8+55.2 261.0+58.3 157.4+34.43 160.1+32.7 7.75+7.2
9.26+0.4 * 124.1+37.3 * 115.6+30.2 * 146.2+41.5 136.2+55.0 * 101.9+34.9 97.1+32.9 * 118.6+39.6 * 119.0+37.1 * 56.5+27.7 57.1+27.2 80.7+42.2 80.3+44.2 29.5+16.4 31.1+12.8 * 134.1 + 39.3 * 151.9 + 45.8 * 235.7 + 45.0 232.3 + 54.3 144.2 + 33.2 146.9 + 30.2 9.26 + 9.8 *
* P 5 0.05.
Table 3. Abbreviations used to identify each variable. Variable
Abbreviation
Mass Dominant femur diameter Dominant limb abduction-with-rotation angle at peak torque at 2.08 rad × s71
mass Dom femur Dom abd-with-rotation angle at peak torque 2.08 rad × s71 Non-dom add-with-rotation PT 0.52 rad × s71 Non-dom abd PT 0.52 rad × s71 Non-dom add-hamstring PT ratio Non-dom add angle PT 3.66 rad × s71 Knee ext PT bilateral deficit 3.14 rad × s71
Non-dominant limb adduction-with-rotation peak torque at 0.52 rad × s71 Non-dominant limb abduction peak torque at 0.52 rad × s71 Non-dominant adduction to hamstring peak torque ratio Non-dominant limb adduction angle of peak torque at 3.66 rad × s71 Knee extensors peak torque bilateral deficit at 3.14 rad × s71
Table 4. Predictor variables that differentiate injured and non-injured players Variable Non-dom add angle PT 3.66 rad × s71 Non-dom abd PT 0.52 rad × s71 Dom abd-with-rotation angle PT 2.08 rad × s71 Mass Dom femur Knee ext PT bilateral deficit 3.14 rad × s71 Non-dom add-with-rotation PT 0.52 rad × s71 Non-dom add-hamstring PT ratio
Wilks’ lambda
F
P-value
0.94932 0.83120 0.75880 0.96107 0.88920 0.99215 0.94429 0.98546
4.8584 18.4800 28.9254 3.6862 11.3394 0.7201 5.3686 1.3422
0.0300 5 0.001 5 0.001 0.0580 0.0011 0.3983 0.0227 0.2497
633
1.00 1.00 0 .44 1.00 0.12 70.099 1.00 0.29 0.07 0.01 1.00 0.39 0 .08 0.29 0.05 1.00 0.13 0 .06 0.01 0.05 70.01 1.00 70.02 0.44 0 .19 0 .09 0.22 0.08 1.00 0.17 0 .001 70.28 0.01 70.06 70.15 70.09 Non-dom add angle PT 3.66 rad × s71 Non-dom abd PT 0.52 rad × s71 Dom abd-with-rotation angle PT 2.08 rad × s71 mass Dom femur Knee ext PT bilateral deficit 3.14 rad × s71 Non-dom add-with-rotation PT 0.52 rad × s71 Non-dominant add- hamstring PT ratio
Knee ext PT Non-dom add-with- Non-dominant Non-dom add Non-dom Dom abd-withangle PT abd PT rotation angle rotation PT add-hamstring Body Dom bilateral deficit 3.66 rad × s71 0.52 rad × s71 PT 2.08 rad × s71 mass femur 3.14 rad × s71 0.52 rad × s71 PT ratio
Table 5. Pooled within-groups correlations among predictor variables for players
A predictive model of groin injuries
predictor variables were low in the majority of cases. The maximum common variance between body mass and peak torque generated during abduction at 0.52 rad × s7 1 using the non-dominant limb was 19.4%. When variables are highly correlated, the variance may be inflated. The eigenvalue for this model was 0.9544. In Table 6, the Wilks’ lambda value was listed as 0.5117, indicating there was a large amount of variability between the non-injured and injured groups but a small amount of variance within the respective sample groups. The chi-square value was 58.297 with eight degrees of freedom. As the observed significance level for the discriminant function was 50.001, the null hypothesis that the means are equal in the non-injured and injured players was rejected. Standardized coefficients take into consideration the simultaneous contributions of all the other variables and indicate the relative contribution of a particular variable to the discriminant function. These coefficients are illustrated in Table 7. Consequently, the large coefficients recorded for non-dominant limb abduction peak torque at 0.52 rad × s7 1 (0.70071) and dominant limb abduction-with-rotation angle at peak torque at 2.08 rad × s7 1 (0.69179) reflect the importance of these two variables. The centroid is the average discriminant score for each injury status group (non-injured and injured groups) and indicates the distance between the two groups. The centroids for the non-injured and injured players were 0.52190 and 7 1.78937, respectively. By substituting the raw score for the discriminating variables into each equation, a player’s discriminant score was calculated. The discriminant score was then compared with the centroid for the non-injured and injured players. Classification into the appropriate injury status group was based on the proximity of the individual’s score to the centroid of the appropriate group. Fisher’s linear discriminant function coefficients provide a direct method of classification. A set of coefficients was obtained for the non-injured and injured group and a player was assigned to the group for which they had the highest discriminant score. These coefficients are displayed in Table 7. The effectiveness of the discriminant function is indicated by the percentage of cases correctly classified. As shown in Table 8, 91.4% of footballers were correctly classified. This model correctly identified 90.5% of players with a predisposition to groin injury and successfully predicted 91.7% of those players who would not sustain a groin injury. This result is substantially better than chance classification (k = 0.79) and indicates that eight variables can be included in a multivariate model that differentiated rugby league players based on predisposition to groin injury.
634
O’Connor
Table 6. Canonical discriminant function values for non-injured and injured players Function
Eigenvalue
Percent of variance
Cumulative percent
Canonical correlation
0.9544
100.00
100.00
0.6988
After Function
Wilks’ lambda
Chi square
d.f.
Significance
0
0.511667
58.297
8
5 0.0001
1
Function 1
Table 7. Standardized canonical discriminant function coefficients and Fisher’s linear discriminant function coefficient for noninjured and injured players Variable
Standardized
Non-injured
Injured
70.519 0.701 0.692 70.581 0.577 70.324 0.178 70.063
0.125 0.005 3.136 70.101 54.152 70.609 0.022 0.077 7310.853
0.224 70.041 2.655 0.039 51.175 70.541 0.015 0.082 7278.913
Non-dom add angle PT 3.66 rad × s71 Non-dom abd PT 0.52 rad × s71 Dom abd-with-rotation angle PT 2.08 rad × s71 mass Dom femur Knee ext PT bilateral deficit 3.14 rad × s71 Non-dom add-with-rotation PT 0.52 rad × s71 Non-dom add-hamstring PT ratio Constant
Table 8. Summary of the classification results for non-injured and injured players Group
No. of cases
Non-injured
72
Injured
21
Predicted group membership Non-injured
Injured
66 91.7% 2 9.5%
6 8.3% 19 90.5%
Note: Percent of ‘grouped’ cases correctly classified = 91.4%.
Discussion The results of this study indicate that the assessment of certain anthropometric and isokinetic variables could aid in the identification of rugby league players who are susceptible to groin injuries. The low Wilks’ lambda values and the pooled within-groups correlations between discriminating variables and canonical discriminant functions, illustrate the importance of three of these variables in differentiating between the injured and non-injured players. These three variables were dominant limb abduction-with-rotation angle at peak
torque at 2.08 rad × s7 1 (0.759), non-dominant limb abduction peak torque at 0.52 rad × s7 1 (0.831) and dominant femur diameter (0.889). It can be speculated that the smaller femur diameter of the injured players may have altered the origin and insertion point of the adductor longus, which may, in turn, alter the efficiency of this muscle and its ability to generate force and cope with stress. However, this explanation is currently unconfirmed and warrants further investigation. The non-injured players generated higher peak torques at 0.52 rad × s7 1 in both hip abduction and adduction-with-rotation than the injured players. In contrast to Knapik et al. (1991) and Lysens et al. (1989), the results of the current study revealed that, regarding absolute strength, the weaker athletes were more likely to be injured. However, Tyler et al. (2001) illustrated that a lack of adductor strength was a factor in the incidence of adductor strains among ice hockey players. Kannus and Jarvinen (1991) indicated the angle of peak torque was influenced by the strength of the appropriate muscle. Consequently, peak torque would be generated later in the range of motion in a weaker muscle, possibly due to slow neural recruitment. In this study, peak torque occurred earlier in the range of motion at 0.52 rad × s7 1 (abduction) and 3.66 rad × s7 1 (adduction) for the non-injured players
635
A predictive model of groin injuries compared with the injured players. Langrana et al. (1984) identified an association between peak torque angle and back pain when assessing trunk flexion and extension. These authors revealed that individuals with back pain generated their peak torque later in the range of motion than symptom-free individuals. Consequently, this model revealed that a player who registers a smaller femur diameter, generates less force in abduction and achieves peak torque later in the abduction-with-rotation movement, is more likely to sustain a groin injury during their playing career. Standardized canonical discriminant function coefficients indicated that the above mentioned variables with the inclusion of body mass are the most important variables in determining groin injury predisposition. Although not discriminators at the univariate level, non-dominant peak torque during adduction-withrotation at 0.52 rad × s7 1, knee extensor bilateral deficit at 3.14 rad × s7 1 and non-dominant adductors to hamstrings peak torque ratio were included in the predictive model and classification accuracy was enhanced. In a broad sense, strength is related to six of the significant variables derived in the predictive model but these variables were not measuring the same construct. This was evident from the pooled within-groups correlation matrix, which indicated that these variables had only low common variance (r = 0.01–0.22), excluding non-dominant adduction to hamstring peak torque ratio and non-dominant limb adduction-with-rotation peak torque at 0.52 rad × s7 1 (r = 0.44) and consequently do not measure the same component because of the independence from one another. This observation indicates that the non-injured players were superior to the groin-injured players in more than one construct and that the correct classification of players to injury status groups relied upon a test battery of eight items rather a single aetiological factor. Interestingly, all but one of the isokinetic variables identified as discriminators involve assessment of the non-dominant limb. The conclusion that groin injuries in rugby league are multirisk phenomena with various risk factors interacting at a given time rather than a single aetiological factor being solely responsible for injury supports the view of Lysens et al. (1991). This model did not find any factors associated with flexibility that were predictive of subsequent groin injury. In support of the results of Tyler et al. (2001) and Emery and Meeuwisse (2001), the screening assessment revealed no consistent patterns between results on individual flexibility tests and the incidence and nature of injuries. The unexpected exclusion of flexibility from the model may be explained by the large variance reported between players regardless of injury status.
The results of this study revealed several differences between non-injured players and those that had sustained a groin injury. The site of the adductor strain (left, right or bilateral) was not influenced by leg dominance; peak torque and power output in adduction and abduction were higher in players who remained injury free; while flexibility scores generally did not differentiate between injured and non-injured rugby league players. Chi-square analysis revealed that playing experience, kicking, side-stepping, level of competition and playing position appeared to have little association with a player’s predisposition to groin injury.
Conclusion The overall findings of this study have allowed the identification of players with a predisposition to groin injury by using a multifactorial model. A large number of kinanthropometric, isokinetic and flexibility variables were measured and discriminant analysis highlighted the importance of just eight of these in determining a player’s susceptibility to groin injury. This illustrates the multifactorial nature of groin injury in contrast to the conventional belief of a single risk factor influencing injury. Therefore, by including these variables in pre-season assessment, players who may have been prevented from playing due to groin injury can be identified. Prophylactic measures could then be taken to reduce the chance of groin injury incidence. Clearly there is a need to determine if, and what forms of, intervention provide the greatest prophylaxis. The variables that made a significant contribution to discrimination between the injury status groups, in order of importance, were: . peak torque of the non-dominant limb during abduction at 0.52 rad × s7 1; . dominant limb abduction-with-rotation angle at peak torque at 2.08 rad × s7 1; . body mass; . femur diameter; . angle of peak torque during adduction at 3.66 rad × s7 1 with the non-dominant limb; . peak torque bilateral deficit of knee extensors at 3.14 rad × s7 1; . non-dominant limb adduction-with-rotation peak torque at 0.52 rad × s7 1; . non-dominant adduction to hamstring peak torque ratio. This model provides insights into the isokinetic and anthropometric constructs deemed important in groin injury predisposition.
636 References Cahalan, T.D., Johnston, M.E., Liu, S. and Chao, E.Y. (1989). Quantitative measurements of hip strength in different age groups. Clinical Orthopaedics, 246, 136–145. Donatelli, R., Catlin, P.A., Backer, G.S., Drane, D.L. and Slater, S. (1991). Isokinetic hip abductor to adductor torque ratio in normals. Isokinetics and Exercise Science, 1, 103–111. Ekstrand, J. (1994). Injuries in soccer: prevention. In Clinical Practice of Sports Injury Prevention and Care (edited by P. Renstrom). pp. 285–293. Oxford: Blackwell Scientific. Emery, C.A. and Meeuwisse, W.H. (2001). Risk factors for groin injuries in hockey. Medicine and Science in Sports and Exercise, 33, 1423–1433. Estwanik, J.J., Sloane, B. and Rosenberg, M.A. (1990). Causes of groin strain. Physician and Sportsmedicine, 18, 55–60, 65. Fried, T. and Lloyd, G. J. (1992). An overview of common soccer injuries: management and prevention. Sports Medicine, 14, 269–275. Gibbs, N. (1993). Injuries in professional rugby league: a three year prospective study of the South Sydney professional rugby league football club. American Journal of Sports Medicine, 21, 696–700. Kannus, P. and Jarvinen, M. (1991). Knee angles of isokinetic peak torque in normal and unstable knee joints. Isokinetics and Exercise Science, 1, 92–98. Knapik, J.J., Bauman, C.L., Jones, B.H., Harris, J.M. and Vaughan, L. (1991). Preseason strength and flexibility imbalances associated with athletic injuries in female collegiate athletes. American Journal of Sports Medicine, 19, 76–81. Kulund, D.N. (1982). The Injured Athlete, pp. 358–359. Philadelphia, PA: J.B. Lipponcott. Kushner, S., Reid, D., Saboe, L. and Penrose, T. (1992). Isokinetic torque values of the hip in professional ballet dancers. Clinical Journal of Sports Medicine, 2, 114–120. Langrana, N.A., Lee, C.K., Alexander, H. and Mayott, C.W. (1984). Quantitative assessment of back strength using isokinetic testing. Spine, 9, 287–290. Lynch, S.A. and Renstrom, P.A. (1999). Groin injuries in sport: treatment strategies. Sports Medicine, 28, 137–144. Lysens, R., Steverlynck, A., van den Auweele, Y., et al. (1984). The predictability of sports injuries. Sports Medicine, 1, 6–10.
O’Connor Lysens, R.J., Ostyn, M.S., van den Auweele, Y., Lefevre, J. and Vuylsteke, M. (1989). The accident-prone and overuse-prone profiles of young athletes. American Journal of Sports Medicine, 17, 612–619. Lysens, R.J., de Weerdt, W. and Nieuwboer, A. (1991). Factors associated with injury proneness. Sports Medicine, 12, 281–289. Meeuwisse, W.H. (1994a). Assessing causation in sport injury: a multifactorial model. Clinical Journal of Sports Medicine, 4, 166–169. Meeuwisse, W.H. (1994b). Athletic injury etiology: distinguishing between interaction and confounding. Clinical Journal of Sports Medicine, 4, 171–175. Merrifield, H.H. and Cowan, R. (1973). Groin strain injuries in ice hockey. Journal of Sports Medicine, 1, 41–42. Morelli, V. and Smith, V. (2001). Groin injuries in athletes. American Family Physician, 64, 1405–1414. Nicholas, S.J. and Tyler, T.F. (2002). Adductor muscle strains in sport. Sports Medicine, 32, 339–344. Norkin, C.C. and White, D.J. (1985). Measurement of Joint Motion: A Guide to Goniometry. Philadelphia, PA: F.A. Davis. Norton, K. and Olds, T (1996). Anthropemetrica. Sydney: University of New South Wales. Poulmedis, P. (1985). Isokinetic maximum torque power of Greek elite soccer players. Journal of Orthopaedic and Sports Physical Therapy, 6, 293–295. Reilly, T. and Florida-James, G. (1995). An analysis of fitness profiles and matchplay demands in Gaelic football. Journal of Sports Sciences, 13, 43. Renstrom, P. (1992). Tendon and muscle injuries in the groin area. Clinics in Sports Medicine, 11, 815–831. Schootman, M., Powell, J.W. and Torner, J.C. (1994). Study designs and potential biases in sports injury research: the case-control study. Sports Medicine, 18, 22– 37. Smodlaka, V.N. (1980). Groin pain in soccer players. Physician and Sportsmedicine, 8, 57–61 Tyler, T.F., Nicholas, S.J., Campbell, R.J. and McHugh, MP. (2001). The association of hip strength and flexibility with the incidence of adductor muscle strains in professional ice hockey players. American Journal of Sports Medicine, 29, 124–128.
